I scroll Instagram like I have midnight snacks: furtively, trying not to disturb the others in my household. The sound is off, the glow of my screen directed away from my husband as I scroll in bed.

We got a puppy in July; Kaylee is a Lab mix, puppy soft, and extraordinarily sweet while she’s not stealing and chewing everything that’s not nailed down. I’ve never had any sort of Lab before, so I started researching everything about them—their habits, their personalities, tips for training them. Of course, my searches followed me to Instagram, where the algorithm began to serve me Labrador content. “This is my 8-week-old Lab’s schedule!” “Look at the tricks my 12-week-old Lab can do!”

During my soundless scrolling, I was served a video from MiamiDogDaddy, who offered homeopathic remedies for typical Lab ailments, along with DIY ways to entertain your dog—have them bob for fruit! Drill holes in a cabbage and fill them with peanut butter!

It wasn’t until MiamiDogDaddy posted a video about combating typical Labrador parasites—while holding up a jar writhing with them—that I paused before scrolling to look more deeply. In the top left corner of the video, in a small font, alternating with the information about the video’s “sound,” was a label: AI Creator.

As we write on Substack today, in our latest investigation that was spawned by my Labrador-obsessed scrolling:

Neither MiamiDogDaddy nor his dogs exist.

[I was] far from the only person who has been misled; over 211,000 other users have followed the account, one of a few dozen “AI creators” affiliated with Iron Paws, a dog supplement brand.

From there, and I combed through the accounts MiamiDogDaddy was following to find how far the network extended; we identified nearly 50 AI-generated accounts promoting the Iron Paws supplement, with millions of collective followers. Using the Meta Content Library, Ben was able to estimate that the network had generated at least 27 million views for Iron Paws under the hashtag #ironpawspartner.

I checked each account to see if Meta labeled them or if they disclosed they were AI-generated. Meta only applied the small “AI Creator” label to 21.3% of accounts we identified, though 61.7% of accounts voluntarily disclosed their AI origins with the hashtag “#syntheticperformer”—an ironic hashtag if there ever was one. Can an AI persona that doesn’t exist, doesn’t have memories, and has never experienced the loss of a pet, “perform?”

In total, over one-third of the accounts—which impersonate breeders, mystics, and even vet techs—had no AI disclosures at all.

I hope you’ll read the rest of the ASP investigation here:

Read more about AI Dog Daddies!

These sorts of AI-generated marketing campaigns are rampant; our investigation follows from reporting by 404 Media, Factchequeado, and The New York Times on AI avatars hawking human health supplements. As AI and deepfake expert Henry Ajder told us, “The ease of which you can spin up these kinds of campaigns now and the relatively low cost and time it takes to do it is why I think we’re seeing much more of it.”

And unfortunately for the billions of people consuming vertical video on social media, they’re pretty convincing. Even I was fooled the first few times I encountered MiamiDogDaddy’s content.

Here are a few tips for recognizing this content—though it’s becoming more difficult by the day:

• Check for the “AI Creator” label at the top left corner of the video if you’re using Instagram on mobile, or under the username if you’re looking on desktop.

• Check for a “#syntheticperformer” or similar hashtag under the video.

• Put the sound on—AI-generated audio can sound less natural and more clipped than normal human speech.

• Look at the background of the video—is text legible? Does it look like a normal home/environment, or is it too perfect?

These aren’t foolproof—as our investigation proved, labeling is not well-enforced, and AI is becoming more convincing every day. My biggest piece of advice (which I should also take!) is to slow down when you’re consuming content online, and as soon as somebody starts selling you something, start asking questions—who is this person? do they exist? what’s their motivation?—and if you can’t answer any of them, don’t buy their supplement for $39.99 a month. 🧭

OVER THE PAST FEW DAYS I’ve been feeling somber and sad as I’ve reflected on the apparent suicide of Jason Arday, a Cambridge University professor who experienced a multi-point societal failure that turned his life upside down and ultimately led to his death.

For those that missed the media frenzy or avoid X.com, where the peanut gallery made Arday’s story trend in the United States, Sangita Myska of The Nerve brings us up to speed:

Jason Arday’s story – as he tells it – was nothing short of extraordinary. He grew up on a council estate in Clapham, south London. [...] Jason was diagnosed with autism, global developmental delay and dyslexia. He claims he was non-verbal until the age of 11, and could not read or write until he was 18. Yet, by 37, he was appointed a professor of sociology at a world-class academic institution. [...]

Now, serious questions have arisen over his backstory. There are allegations that he heavily embellished his CV. His claim that he was non-verbal at primary school has been undermined by a former classmate who says Jason Arday was so talkative as a child that his nickname was “chatty-patty”. Then there were his undocumented descriptions of racial harassment and stalking. Significantly more serious than all that were the accusations of plagiarism – that he stole other people’s research without sufficient attribution, change or development. Arday has admitted to “mistakes” in his PhD but denies the claim of plagiarism.

But, as writes:

The allegations trace back to Liverpool John Moores University, which awarded Arday’s doctorate, previously investigated plagiarism claims, and did not uphold them. The latest campaign was reignited in part by Nathan Cofnas, a prominent anti-DEI academic who published his accusations under the headline “DEI Fraud and Cover-Up at Cambridge.”

Cofnas, by the way, published his claims on this platform (Substack). Karen is being charitable calling him an “anti-DEI academic;” he calls himself a “race realist” and was fired from Cambridge for asserting that there would be “close to zero” Black professors at Harvard if it were a meritocracy. Back to Attiah:

Arday resigned on August 5th from his dream position at Cambridge, citing that “relentless accusations, speculation and public commentary” had taken a “profound toll” on him.

Nine days later, he was dead.

Arday’s story has brought up some serious questions about the academy and about publishing. For instance: Did Cambridge fully vet Arday before they hired him, or did they set him up to fail? Why, in this day and age, did Simon and Schuster not subject Arday’s memoir, which the carrion crows of the internet picked apart over the past few weeks, to a thorough legal and fact-checking read to protect both their author and their investment?

However—and as usual—this trumped-up scandal wasn’t focused on serious inquiry or making real change. It was about hounding a person out of their job and ruining their life—and in this case, tragically, ending it.

Jason Arday’s story shows that the right wing’s targets can do everything they demand—resign, admit to wrongdoing—and it still won’t satisfy the rabid online mobs.

Not even Arday’s death will.

Instead, they are moving on to other targets they claim enabled Arday, while continuing to malign and ridicule a 41-year-old father who can no longer defend himself—and his family. Yesterday, I was sent a post from a member of the UK’s right-wing reform party that ridiculed the fundraiser as a “family grift.” Arday left behind a wife and two children, and the mobs who targeted him would begrudge them financial support equal to a few years’ of his salary.

Subscribe now

wrote eloquently about mainstream and social media’s role in Arday’s story. As she points out, after Arday’s resignation, and his request to be left alone, “the media hounding didn’t just continue, it increased,” with some outlets buying ads to amplify their grotesqueries masquerading as “opinion” and “investigation.”

“The cultivation of targets that can be righteously attacked is the business model,” Carole writes. “This is what surveillance capitalism has given us. A recursive feedback loop that sees an ad team select a publication’s most clickbaity articles that leads to an increase in traffic that signals the “public interest” aka appetite to publish more articles. It’s a money machine.”

Carole also ends her piece with a list of ways to support targets of harassment. I shared the following with her:

I worry about the many people who will undoubtedly make themselves smaller after this campaign of hate. The people of color, the women who will eschew a life that has any public component for fear of what might be unleashed on them. Even if dismal human beings like [Nathan] Cofnas don’t succeed in literally policing academia or publishing, they may succeed in their campaign of self-censorship, unless institutions develop much more robust support structure for targets of networked harassment campaigns like this.

It’s those support structures—providing mental health support; assisting those who have been doxxed move to a second location; proactive monitoring and communications that don’t just protect the institution, but the individual target—those are the things that can bring people back from the brink.

A lot of institutions still subscribe to the idea that “keeping your head down” and “waiting for the scandal to blow over” are part of a winning strategy in the internet age. They couldn’t be more wrong.

Despite his resignation, the right-wing, racist, bloodthirsty mobs of the internet and the media sector that encouraged them would have continued to attack Jason Arday, were he still alive.

I know, because more than four years after experiencing my own digital witch hunt, my detractors still regularly find reasons to bring me up, to malign me, and to wish violence upon me and my family. Just this week, a “satire” news site (that deeply buries this label) claimed that I had died after being imprisoned in GITMO. Apparently this passes as a joke in MAGA land. As of this morning, two days after publication, the “article” has more than 30,000 views. It has hundreds of comments from people who clearly haven’t all read the site’s disclaimer about its content—one particularly nasty example reads “HANG HER DEAD CORPSE NOW !!!!! 🔪🔪🔪🔪🔪.” (This “parody” site has also raised over $400,000—where are the right wingers aghast at the grift?)

Study after study has established that people with intersectional identities (a fancy way of saying people who represent multiple minorities, like Arday, who was Black and autistic) experience online abuse at higher quantities and levels of severity. (Here’s a very fresh study from University College London that demonstrates this phenomenon.) We’ve known this for years—the first study I did that made this conclusion was published in 2020, and I was far from the first person to reach it.

Yet here we sit, in 2026, shocked at what antiquated institutions, the media, and the internet have wrought. We should not be.

There will be postmortems, investigations, further speculation and ridicule. Some institutions may publish lessons learned and make feeble, bureaucratic attempts at protecting their people, while others will have learned a completely different lesson: that stories like Arday’s are moneymakers, humanity be damned.

In all but the most extreme of cases, public harassment campaigns have become normalized, viewed as the “cost of doing business” or the price of speaking out. A target who raises their voice in opposition to their public haranguing is often met with derision—“if you can’t take the heat, get out of the kitchen”—or a claim they deserve what they’re experiencing for whatever crime they’re perceived to have committed.

And that’s the core of this tragedy. Multiple systems failed Jason Arday, but perhaps none so much as our common humanity.

We’ve been conditioned to wonder whether a person has done enough to deserve compassion, but when we’re asking that question, we’ve already lost. 🧭

Worth reading in full:

• Sangita Myska: Is Jason Arday a fantasist or a victim of racism? Actually, he’s both

• Karen Attiah: Jason Arday’s Death and the Age of Digital Racial Terror

• Carole Cadwalladr: A Digital Lynching

• A tribute from one of Arday’s friends, George Ryan

LAST WEEK, far-right influencer Mike Cernovich tweeted that I was nearly “given complete control over the internet” in my former role in the Biden Administration.

If you don’t know Cernovich, he’s a manosphere personality who rose to prominence during the #Gamergate harassment campaign and later cashed in on the Pizzagate conspiracy theory. The Southern Poverty Law Center writes that he “made his career on trolling the liberal establishment by accusing people of pedophilia or child sex trafficking. Armed with a law degree, Cernovich claims to defend ‘free speech,’ in particular the freedom to harass women and make misogynistic, violent comments.”

(That tracks; after he tweeted about me, X.com morphed into a Throwback Thursday of four-year-old lies about me, along with requisite harassment and threats. A prominent conspiracy theorist publicly called for my swatting for purported ongoing crimes.)

Now, Cernovich is cashing in on the MAHA movement, at least according to his Instagram bio: “Filmmaker, investor, and e-commerce co-founder of global brands. Currently launching a protein bar.”

Super nice guy. Obviously a paragon for trustworthy information.

To be clear: I never came close to having any sort of control over the internet. (You can learn about the boring truth of what my job in government was here.) I didn’t want that control. No one should have it. And any job that included any power to adjudicate speech is one I would have publicly condemned, in line with my career of anti-censorship research and advocacy.

But since Mike Cernovich is fantasizing about me, I thought it might be fun to do a little fantasizing of my own.

Here are five common sense policies I would pursue if I had “complete control” of the internet. I think the monetized liars might be surprised.

Give Americans the right to file appeals against social media company platform moderation decisions. Did you know that social media users in Europe can file appeals when they believe social media companies have taken down their content erroneously? Since this rule came into force, “30% of 165 million content moderation decisions that users appealed through the platforms’ internal mechanisms have been reversed.” Europeans have more online rights than Americans, despite what the monetized liars may say.

Get researchers access to platform data. Europe also has enshrined access to platform data for researchers and journalists in its internet regulation. This allows civil society to hold platforms and governments to account. Allegations of censorship or “complete control” of the internet, for instance, could be scrutinized through this data. Without it, conversations about what’s happening online amount to a very tiresome he-said-she-said. Who are we protecting if we are blocking measures like this? It’s certainly not citizens.

Dismantle the infrastructure behind deepfake nudification. Many countries—including the United States, in the context of the TAKE IT DOWN Act, which became law this year—are imposing penalties on those who create, share, or enable the sharing of deepfake non-consensual intimate imagery. (Earlier this year, the Senate also passed the DEFIANCE Act, which would create a civil right of action for victims of deepfake abuse; it is awaiting action in the House.)

But what about the infrastructure that allows the images to be created in the first place—the servers that host websites dedicated to deepfake websites and nudifying apps? The app stores that serve them to users? The payment processors that facilitate the subscriptions or the exchange of money for the creation of a deepfake? If every part of the deepfake infrastructure were forced to implement Know Your Customer procedures the way banks are forced to vet potential clients for signs of terrorism and money laundering, the problem would not be the worldwide scourge it is today.

Make platform-level AI opt-in—not opt-out—the default. Some folks don’t want to read an AI-generated knowledge panel. They would rather avoid an AI-generated AI detector (looking at you, Substack). They would prefer their emails and documents stay their own, not hoovered up by an app’s AI model for training purposes. Rather than forcing users to figure out how to opt out of opaque systems, platforms should make them opt in by default, seeking our informed consent for both using our data in training and turning on AI features on platforms.

Incentivize on-platform information literacy education. I’ve been harping on this for years. The multi-billion dollar platforms that are poisoning our information ecosystem should educate users about the manipulative features of their platforms. They have ubiquitous access to millions of people. They encourage people to register to vote and cast a ballot, but they don’t educate them about why and how information impacting their vote made its way to them, because it impacts their bottom line. What would it look like if platforms ran compelling public service announcements about AI-generated scams and fraud? The basics of algorithmic amplification?

Those are the first five things that popped into my head. Notice how none of these involve anything even remotely related to the removal of political speech! A power-hungry would-be censor would be focused on that, and we don’t have to speculate what it might look like. We have hundreds of examples from the current administration that has been cataloguing for the past year and a half over at TrumpCensorship.com.

Just sayin’. 🧭

Online Notes

A few pieces of news and analysis from recent weeks

LinkedIn Adds a “Seems Like AI Slop” Button

Last week, I wrote about Substack’s decision to partner with Pangram, an AI-powered AI detector (yes, you read that right), in order to give users more visibility into how content on this platform is being created. In Substack’s rollout materials on the partnership, Co-Founder and CEO Chris Best took several opportunities to dunk on LinkedIn. “We don’t want to wait until your Substack app turns into LinkedIn before we start to learn and make progress,” he all but chortled.

LinkedIn answered that potshot this week with an announcement from its Chief Product Officer, Hari Srinivasan that the platform had rolled out a “Seems Like AI Slop” button. In an opaque reference to Substack, he wrote: “We want members to get feedback from real humans on what sounds authentic - not just have an AI detector review it and get it wrong.”

While I think this is closer to the right solution than publicly assigning an AI-generated scarlet letter to any post that Pangram thinks deserves it, LinkedIn’s effort also leaves something to be desired. I would have liked to have seen a little more information from the platform on what happens next. At least from Srinivasan’s post, it seems like right now this button is just generating signals for the platform to interpret and decide what content moderation actions to take, and that, like Substack, LinkedIn is making it up as it goes along.

Mozilla Audited Invasive Cycle Tracking Apps—Here are Two Alternatives

The Mozilla Foundation published a useful audit of period tracking apps. As they write:

Period tracking apps aren’t just about when your period happens. Apps in this space ask you to log extremely intimate details so it can predict your cycle or help you get pregnant. From when you bleed, but also when you’ve had that late-night chocolate craving, and even when you have a miscarriage. In the Post-Dobbs landscape, we believe safeguarding this data and keeping it private is more important than ever.

If you or a woman in your life use a cycle tracking app, it’s worth switching to an app that takes privacy seriously. Here’s how Mozilla rates a few popular apps:

A privacy-minded friend also recommended Drip to me; I’ve used it for years and can confirm it checks all the boxes for privacy and is incredibly user-friendly.

The Horrifying World of Spy Cam Voyeurism

A couple of years ago, a friend was visiting Washington and was a few nights into an Airbnb stay in Georgetown when he made a horrifying discovery: his host had installed cameras throughout the property. He covered them, understandably creeped out by the potential for his host to spy on him.

The host wasn’t happy; he kicked my friend out the next morning. Airbnb has since updated their policies to globally ban this sort of spying. But imagine if the footage of my friend showering, using the bathroom, or any other private activity had been uploaded to the internet. According to a new BBC documentary presented by , there’s a growing market both for this voyeurism and the tools necessary to create it.

I was lucky enough to meet Jess, an advocate for women’s rights and privacy and author of No One Wants to See your D*ck, in 2024 when the Department of Health and Human Services held an event on ending image-based sexual abuse. I was impressed by her conviction and have followed her ever since. Her new documentary explores the amateur forums that equips men to spy on women in hotels, dressing rooms, restrooms, and beyond; Jess even confronts one of the site owners, who whines that “men are unhappy” about laws like the UK Online Safety Act, which are trying to root out behavior like this. It’s well worth 30 minutes of your life.

In case you missed my latest…

This week I wrote about why Sarah Rogers, Trump’s pick for the CEO of the US Agency for Global Media, which oversees Voice of America and Radio Free Europe, among other foreign broadcasting properties, is just a MAGA monster in free speech clothing:

Offline Notes

The real-life pastimes, products, and programs getting me through the infopocalypse.

As I’ve written before, I’ve been spending a lot of time in the garden this summer. Earlier this year, as I was researching how to deal with the hard clay soil that is typical for my region, I heard a great conversation on NPR’s 1A about native plants and their role in supporting threatened species. Natives can also thrive in poor soil conditions without the construction of raised beds or making amendments to the soil. Natives, I decided, were the way to go.

I spent five years working in a plant nursery as a teenager, but I didn’t know the first thing about natives. Luckily, my very first boss in Washington, , who had a long career in democracy support, is now a native plant landscape specialist. She created a gorgeous plan for us and we’ve been enjoying caring for our new plants throughout the summer.

Lately, the biggest burden has been weeding. We had a few rainy weeks in a row, followed by the oppressive smoke cloud from the Canadian wildfire. I was heads down on some big projects, and before I knew it, I had some massive crabgrass specimens snaking their tentacles through my beds. It took a few hours of hard work over two weeks to claw them back. I finished last weekend before we went out of town, and by the time I was out in the garden to water this week, the first bed I finished had more weeds peeking through again.

While I’ve been weeding and watering, I’ve enjoyed listening to a book I’ve come to think of as writer church: Susan Orlean’s Joyride, a recounting of The New Yorker staff writer and The Orchid Thief author’s life and career, and, more importantly (for me anyway!) a peek inside her process. I’m glad I am not the only writer walking into interviews with very little idea what I’m going to be asking my subjects about! 😅

Enjoy the rest of your weekend! If you enjoyed this post, send it to a friend or hit the heart button. 🩵

This email contains affiliate links to Amazon.com, whereby I earn a small commission on any purchases you make.

PICTURE, IF YOU WILL, a person who has cheered the both barring of researchers from the United States based on their speech and the dismantling of an office that fought foreign influence.

A political appointee who appeared alongside an extreme UK politician who pushed “progressive eugenics”—and considered giving groups affiliated with him US government funding.

A person who likened government action against the non-consensual deepfake nudification of women and girls to a “Russia-style ban.”

Someone so hell bent on owning the libs they instructed the State Department to build an internet portal that would bypass other countries’ internet laws and allow users to see illegal content, including neo-Nazi content and hate speech.

That person is Sarah Rogers, who currently serves as the Trump Administration’s Under Secretary of State for Public Diplomacy. Though she has been described in multiple outlets as a “free speech crusader” (see here, here, and here), her actions tell a different story: she is on a free speech crusade for MAGA and its far-right allies alone, not for those who criticize them. She has also been nominated to lead the US Agency for Global Media (USAGM); the Senate Foreign Relations Committee is holding her confirmation hearing this morning.

If confirmed, Rogers—who is a lawyer by training and has no experience running a media organization—would continue her work overseeing America’s rapidly deteriorating public image around the world while bringing her far-right views to the agency that oversees Voice of America (VOA), Radio Free Europe, Radio Free Asia, and other public media properties that are meant to bring reputable information to closed societies.

Subscribe now

USAGM has been decimated under Trump 2.0. In March 2025, election denier Kari Lake—who served as Acting CEO of the Agency while it was without a Senate-confirmed appointee—worked with DOGE to slash a workforce of over 2000 journalists around the world to the statutory minimum of 81. Broadcasts to places like Iran, Venezuela, and Russia—places you might think the Trump administration had an interest in undermining local propaganda—were halted.

The U.S. District Court for the District of Columbia ruled a year later that Lake had illegally served as Acting CEO and that her actions in that position were null and void. The judge ordered USAGM to return former employees to their positions and resume broadcasts.

In the months since, as the Trump administration has fought an unpopular war in Iran, it has attempted to turn USAGM back on—and turn it into a MAGA propaganda outlet. Patsy Widakuswara, former VOA White House Bureau Chief and lead plaintiff against the administration’s attempt to dismantle the agency describes the partisanship that has come to characterize VOA for ’s Substack:

Current editorial priorities bear many of the characteristics of wartime propaganda, including at VOA Farsi, one of only six of VOA’s 49 language services brought back in a much-reduced capacity following legal action against the administration. Stories center on administration statements, often without opposing context. Recent articles highlight administration claims that oil prices are lower than before the war and that maritime traffic through the Strait of Hormuz has been restored. Meanwhile, VOA suppressed coverage of civilian death tolls from U.S. airstrikes and barely mentioned the February bombing of the elementary school in Minab.

The USAGM that Rogers stands to inherit bears no resemblance to the promise Voice of America made to its listeners during its first foreign broadcast on February 1, 1942: “Daily at this time, we shall speak to you about America...The news may be good or bad. We shall tell you the truth.”

Rogers will not do that. Whatever editorial independence USAGM still clings to will be obliterated. Critical coverage of Trump will be softened or canned, as will coverage of his allies such as Nigel Farage or other far-right figures across Europe. She will calcify the transformation of America’s foreign broadcasters into propaganda outlets and continue to squander the trust of audiences that have depended on them for reputable reporting for decades.

America’s advantage in the information war has never been peddling propaganda; it has been the dependable truth. We don’t serve America’s interests by pushing a puffed up, perfidious picture of our country. But our adversaries—whose propaganda is slicker, more compelling, and more practiced than MAGA’s blunt, cringeworthy attempts—will surely celebrate when we do.

I’d tell you to picture it, but you don’t have to. It’s already happening. 🧭

Further reading:

• Just published today: The New Yorker on “How Trump Silenced the Voice of America.”

• in The Atlantic on Kari Lake’s tenure at USAGM.

A COUPLE OF MONTHS AGO, a Google Alert let me know my name was mentioned in a Substack post from a user I hadn’t heard of. The “resistance”-themed account specialized in writing songs and poems and had mentioned me in one of its non-rhyming posts. It got basic facts about my government tenure wrong and made odd conclusions, so I dug deeper. The account posted new content almost daily. Its “poetry” was too long and verged on nonsensical. The pace and the quality made it pretty clear the account was using AI to generate its content.

To my dismay, the account had tens of thousands of subscribers, was monetized, and seemed to be cashing in on the left’s very real desire to support something—anything!—that was taking a stand against Trump. Meanwhile, my own newsletter and others’ sputtered despite putting in real work to write human-generated content on Substack’s so-called “engine of culture.”

I don’t want to read AI slop on Substack—or any other platform where I’m spending hours of my one wild and precious life. I think AI is wildly misogynist—and others have proven it is racist and biased in many other ways that reflect the state of our world. And I’m worried enough about the effects widespread AI use is having on cognitive function that I ban AI use in the graduate-level course I teach.

So I wasn’t necessarily predisposed to dislike Substack’s new partnership with AI-detection tool Pangram. But the tool’s imperfections, along with the platform’s perplexing rollout strategy for the partnership, makes me think this is little more than a vanity project that will allow Substack to claim it is the last human-dominated social media platform.

The Pangram collaboration, which Substack CEO and Co-Founder Chris Best rolled out in The Substack Post on July 21, allows anyone to scan any post, note, or comment on Substack to see if it is fully AI-drafted, AI-assisted, or human-written.

Some writers seem perfectly happy to shrug as an AI tool—itself trained on the stolen works of writers—passes its imperfect judgement on whether our work is also AI-generated. Even though Pangram is widely recognized as best-in-class among AI-detectors, its still makes plenty of mistakes, as writes:

Given that Pangram still gets a false positive 1 out of every 10,000 analyzed posts (and as notes, that’s in controlled lab conditions!), if there are say, 2 million posts a week, that’s 20,000 cases where someone’s content is labeled 100% AI-generated.

Pangram performs worse in the other direction; The Atlantic reported that “Pangram’s false-negative rate, or how frequently the model incorrectly labels text as human, is closer to one-in-70 (although some other assessments say it is more accurate than that).” And after Indicator’s Alexios Mantzarlis conducted his own adversarial test on the tool, he found it fairly simple to get the detector to produce false negatives by making text rhyme or inverting phrases.

Aside from these concerns about accuracy, the way Substack rolled out the feature—and with whom—gives me tremendous pause. Its landmark announcements—the aforementioned piece from Best and a conversation with noted tech bro shill Matt Taibbi—did not address some of the biggest questions users were bound to have about the tool, including:

If I or others scan my text, is it being used to train Pangram’s models? After a lot of folks raised this question, Substack clarified in a post published three days after the launch: “These features do not use publisher content, either through Substack or Pangram, to train generative AI models.”

Is it possible to opt out of this feature? Yes, you can do this on a by-post, by-note basis, only after you scan your own post for AI usage. Substack only introduced this option after user feedback. I worry, though, that if allowing the scans becomes the norm, turning off this feature might be viewed as an admission of guilt. (I’m frankly not sure if I’ll turn it off yet—you’ll find out when I do!)

Further, though users are asking for it, Substack has not allowed users to opt out on a platform level, which seems extremely suspicious particularly when paired with the “first scan, then opt out” flow above. Is the workflow constructed like this because Pangram wants writers to be incentivized to use this “free” tool? What were the terms of the deal? Is Substack paying Pangram on a per-scan basis, or have they paid a flat fee and want to get their money’s worth? Do they want to generate as much data as possible so they can claim they have the “most human-generated content on the internet?”

If your text is marked as AI, does it affect your distribution or ability to monetize content at all? Substack hasn’t really addressed this yet. In Best’s conversation with Taibbi, he says, “Whatever we do I never want Substack to be a cop” and hopes the tool will simply allow users to ascertain if “there’s a trust that’s being violated.”

What about when Pangram gets it wrong? Well, according to the Substack help page about the Pangram collaboration, you, the user, “report [the] detection error,” “enter feedback,” and “it will be reviewed to improve detection quality.” Will the score be reversed? Will it be removed? Will you be paid for the tool’s defamation of your brand, your work, your name? 🤷🏻‍♀️ We’re left to guess.

And, the most telling question for me: Why did Substack choose Matt Taibbi to flog this rollout? Is Taibbi—a man who has spent the better part of the last four years peddling nonsense about disinformation researchers, willfully misinterpreting quotes, and failing to grasp linear time—who they think the most respected journalist in the space is? Or were they just focused on picking someone who would act as a rubber stamp, ask no hard questions, and wave away inconvenient truths? (For example, in the interview with Best, Taibbi excused Pangram’s error rates by noting that the Pangram guys admit that their tool isn’t foolproof. Gee, I’m sure that’s comforting to a person whose career is being lit on fire by a false accusation of AI-generated writing!) Or perhaps most chillingly, Substack leadership could have been interested in getting ahead of the business implications if Taibbi and his merry band of memelords decided that these labels were somehow “censorship”.

It’s frankly shocking that people like Taibbi don’t have more of a problem with the forced deployment of Pangram to every Substack user. The same set of people who cried “censorship” about fact-checking and disinformation research and tools that rate and label websites for trustworthiness are perfectly fine with a machine running an opaque, probabilistic model being used for deterministic conclusions that could affect people’s reputations and livelihoods. Pangram, they assert, is “designed to raise awareness,” but the same is true of the suite of tools and programs that have been deemed Orwellian. The fact is that Pangram performs a role very similar to Facebook’s maligned fact checkers, except it does it more opaquely.

I don’t hate the idea of a tool like this in general—I was and remain a fan of adding fact-checks and context cards to tweets and Facebook posts, and those efforts slowed the spread of misleading information. But efforts like this need to be rolled out with adequate transparency and must include proper appeals processes. With a tool like Pangram, users should also be educated about its limitations.

Between the rollout opacity and their choice of messenger—a man who uses his brain to write, sure, but uses it to write fantasy under the label of “journalism”—I think this is more of a vanity project than an attempt to safeguard humanity or reality. 🧭

Online Notes

A few pieces of news and analysis from the past week

We Should View Trump’s Speech as Preparation for Election Interference

This week, I was happy to contribute a piece to MSNOW’s Opinion Page about President Trump’s desperate, conspiratorial speech this week.

In order to distract from his flagging poll numbers, high gas prices, and the ICE killings that for some reason keep happening, Trump scheduled a primetime address, returning to

a favorite topic: conspiracy theories about the 2020 presidential election. The president made sweeping claims about a broad Chinese campaign to access voter records, agitate against his candidacy and interfere in the election. But the claims are not borne out by the evidence, including the intelligence documents he declassified as “proof.” The speech was a desperate attempt to undermine confidence in our democratic system and lay the groundwork for an unprecedented and illegal attempt to control elections from the White House itself.

In the piece, I provide some context about elections. Trump’s misleading claim that China has access to 200 million Americans’ voter information, for instance, is a lot less alarming when you realize this information is either public or available for purchase on the open internet. I also discuss the background of the documents that Trump declassified:

[They] don’t show some conspiracy by the intelligence community to hide information from the president. They show intelligence professionals doing their jobs: assessing sources, debating the reliability of reports and making judgment calls about when it is necessary to publicly sound an alarm. Disagreement is not a sign of a cover-up, as the president alleged, but a hallmark of intelligence professionals doing their jobs. Disclosing every detected foreign influence campaign, no matter how minor, would risk seriously undermining the American people’s confidence in our elections — exactly as the president did this week.

The documents do, however, leave unredacted the names of two since-fired intel officials that Trump believes were involved in the “suppression” of intelligence about Chinese influence campaigns against him. The documents show these officials recalled two reports because of flimsy sourcing (one source hadn’t provided confirmed information in over a year; the other was a secondhand source). Leaving these names unredacted isn’t an accident. The Trump administration knows exactly what follows: online and offline harassment and threats.

READ THE REST

We Don’t Need an AI Court or AI Journalist Rankings

While in the car this week, I heard an interview on NPR’s Here and Now with journalist Gary Baum, who became the first target of a bizarre, Peter-Thiel-backed AI startup called “Objection AI.” In reporting about the experience, Baum explains:

“Someone has filed an objection against something you wrote,” explained Austin Livingston, pointing me to a web page where Purdue Pharma heir Michael Sackler, a film financier and self-styled ethical investor, had paid a new tech startup — fittingly called Objection — to assess the legitimacy of a skeptical article I’d published about him and his business in The Hollywood Reporter five years earlier.

Livingston, a young staffer who describes his job on LinkedIn as, simply, “creating shareholder value,” noted that an AI tribunal would attempt to “adjudicate a determination of truth,” and that the announced outcome “will also affect your Honor Index score, a measure of the veracity of your published work.” He went on, adding, “You can argue your side by uploading your evidence and suggesting your interpretation of the allegation.”

What dystopian nightmare is this? AI is regularly hallucinating in and out of the legal world, but some people are so high on their own supply they think it can serve as a neutral and truthful adjudicator Baum’s reporting, which is worth reading in full, seems to have convinced the leaders of “Objection” it was a bad idea; it is now an AI-powered journalist-ranking site. To be clear, we don’t need that either.

Offline Notes

The real-life pastimes, products, and programs getting me through the infopocalypse.

I was in New York this week, and got to spend time with lovely friends and colleagues like and Michelle Kinney of the Mayday Network (have you listened to our Americans rewatch podcast, The Illegals?) and . I also ate pizza.

Kaylee, our pup, seems to be getting bigger every day. I sent in a doggie DNA test for her last week and can’t wait to see what breeds other than lab are contributing to her extreme cuteness.

I’ve been spending more time on Instagram, spending a lot of time making more vertical videos and post more analysis on that platform, only to see middling growth. Are you on Instagram? Do you follow me there?

Follow me on Instagram

Hope you enjoy your weekend—I’ll be back in your inbox early next week. 🧭

THE IRRESPONSIBLE SPECULATION around the death of Lindsey Graham and the long, secret illness of Mitch McConnell have provided ample evidence for a fact that may upset some of you: batshit conspiracy theories are not the sole terrain of the Republican party.

As far back as 2018, when I testified in the Senate about countering foreign disinformation, I’ve always repeated a single phrase: “disinformation knows no political party.” In that context and most others, it was meant as a warning to Republicans: you may be inclined to ignore, or even welcome Russian influence campaigns when they help you, but when one day an adversary targets your party, and your candidate, you’re going to regret it. (I was proven right—see Iran’s lego propaganda.)

But another warning—one that is too infrequently acknowledged—was baked into that statement. It’s not just Republicans who are susceptible to disinformation; Democrats can and do fall victim to it, too. In How to Lose the Information War, I told the story of a group of left-wing activists who organized a Les Mis-themed flash mob on July 4, 2017, and were unknowingly collaborating with a troll from Russia’s infamous Internet Research Agency, who had helped them purchase Facebook ads to advertise the event and drive up turnout:

The $80 spent to drive show tune-loving, progressive, D.C.-area residents to the White House to protest the president in song on July 4, 2017, wasn’t a mistake; Russia has long tried to increase discord in American society.

[Disinformation] feeds on, amplifies and weaponizes our emotions, pitting us against one another. It’s not just widespread digital literacy campaigns targeted at voters and social media users we need. Politicians need to sign on as well, calling out disinformation—whether foreign or domestic, no matter which political party it opportunistically supports—as the equal-opportunity threat to democracy it is. If our elected officials and political organizations traffic in disinformation, we are doing Russia’s work for it.

Well, I am sure the Russians who work on influence operations were absolutely delighted when troves of Americans on both sides of the political spectrum immediately worked themselves up into a conspiratorial froth over the death of South Carolina Senator Lindsey Graham on Sunday. Most prominently, those claims started on the right with Laura Loomer, who floated that Russia or Iran might have orchestrated a hit on Graham:

But it wasn’t very long before the conspiracies hopped the gulf between right and left like COVID at a choir practice. Progressives were telling off-color jokes about Graham in one post, and in the next expressing their deep concern that the White House was covering up his poisoning, as if poison was the sole reason a slightly-overweight 71-year-old man who just undertook an arduous journey to a war zone might suddenly drop dead.

In the evening, the DC medical examiner released a statement with their preliminary findings:

Graham suffered from an aortic dissection, in which a tear occurs in the inner layer of the main artery. This was caused by arteriosclerotic cardiovascular disease, the statement said. These tears usually occur when there is high blood pressure.

Within the same hour, Senator Mitch McConnell, who had been the focus of intense speculation and conspiracy theories after his team had suppressed information about his health since he was hospitalized on June 14, released a “proof-of-life” photo, probably in hopes of getting self-appointed investigative journalists to stop creeping around in the back alley behind his house. (Yes, that really happened, but I’m not linking the video because I don’t want to amplify behavior like that.) The photo may have been edited a bit, or else compressed—McConnell’s finger, holding the Saturday edition of the Washington Post Sports section, has a strange halo around it. The edge of the left side of his pillow looks a bit off. There are completely legitimate explanations for why this might be.

But the left-wing internet saw only conspiracy. Almost every single post in my Threads feed on Sunday night was a would-be sleuth sharing their smoking gun: AI had told them the photo was actually from 2023! Or AI-generated! Or legitimate, but enhanced, so the Senator’s seemed younger and wrinkle-free! (Deepfake detection expert Hany Farid posted on Monday that he can find no evidence the photo is AI-generated.)

This all made me pretty distraught. A lot of the prominent folks who were fanning the flames of these conspiracy theories should have known better. The general population was primed for this, between a Senate office that hadn’t communicated in a month and a strong and understandable degree of distrust toward anything anyone allied with the Trump administration said or did. And everyone was using technology that was appealing to their worst instincts—they wanted answers, they wanted the answers to make them feel smart or righteous, and they wanted them now.

As the conspiracies were spiraling out of control, I posted the following two exhortations to reason on my social platforms:

In response, a lot of angry people told me I was being a scold, that I was admonishing them for asking questions, or that I was insane because the McConnell photo was—to their very expert eyes—so obviously manipulated.

I want to be clear: I’m not telling anyone that they should stop asking questions or holding the Trump administration to account. But it’s important to recognize that the behavior we saw on Sunday night—breathless speculation entirely devoid of evidence—isn’t what accountability looks like, and it isn’t accomplishing much besides making social media platforms money. It was incentivized by social media platforms that want to keep us scrolling and seething, generating revenue through advertisements. They created algorithms that reward enragement with engagement, and creators—who rely on social media to pay their bills—know what it takes to get and maintain attention.

What I’m asking of you, and of everyone who aspires to return to a shared reality, is actually a lot more difficult: to actively work against the way we’ve been conditioned by unscrupulous politicians, influencers, and the social media platforms that made them. To actively work against human nature. To practice patience.

These are four maxims that I hope will help you in that mission. I’d also encourage you to check out @ Michael Caulfield’s SIFT Method and The News Literacy Project’s Breaking News Checklist.

1. If the content you’re consuming is salacious, shocking, enraging, or otherwise making you emotional, someone may be manipulating you. We know platforms amplify emotionally-charged content. We know users—especially those who monetize their content—know this. Breaking news events are ripe for engagement bait. If the content were on the front page of the National Enquirer at the supermarket—or better yet, if you saw it on Laura Loomer’s feed—would you believe it?

2. If news is fresh, details will be scarce. News outlets and experts are doing due diligence to verify accounts from human sources and confirm video and photos are authentic. None of that happens instantaneously. For those that make claims with certainty in the hours after news breaks, see #1.

3. AI hallucinates. It can’t do what you’re asking. It probably can’t answer your questions about breaking news, and it definitely can’t accurately verify the provenance of a photo or video. Treat it like an intern that favors speed over accuracy or a student who definitely hasn’t done the reading but keeps answering the teacher’s questions anyway.

4. The simplest explanation is probably the right one. Yes, even now. An aging man had heart problems. A frail man who probably shouldn’t be a Senator anymore did a poor job communicating about his frailty. There are many evil things happening, but they are still outnumbered by plainly stupid choices and unfortunate facts of life.

If you’re a chronically-online engaged citizen who wants to keep up with the news and keep the pressure on the government, you can be a responsible information consumer in a world of AI slop, but you need to start by slowing down and recognizing that you can’t know everything immediately. Your advocacy and your rage is better served when it is grounded in reality. 🧭

Subscribe now

Your feedback matters to me! Please take this short survey to let me know how you’ve enjoyed The Wayfinder over the past six months. 🧭

Take survey

THINK BACK TO SUMMER 2024, to the lime green, meme-laden, hopium-laced “brat summer” before Donald Trump was elected to the presidency a second time. The presidential election was described as the first “influencer election,” when both parties were embracing the power of content creators to inform voters and drive behavior change. The Democrats boosted the clip economy and credentialed over 200 influencers to attend the DNC, among other strategies. But their play looked timid in comparison to the GOP’s long-term gambit, which involved a well-oiled, targeted, free-flowing creation and amplification machine that indexed on longform content.

There’s been a lot of ink spilled on where the Dems went wrong in 2024, as well as the party’s continued struggles building a content machine to rival the GOP’s today. I’m not invoking the ghost of brat summer to discuss which party is better at internetting—the answer is obvious. Instead, I’m thinking about 2024 and its aftermath because it demonstrated the troubling gulf between the mainstream, credentialed press and creators in stark relief.

Take the 2024 DNC, where the Party’s wining and dining of content creators left the mainstream press (rightly) pissed off with their comparatively shabby treatment. Or the ways since then that the mainstream media have irked creators, sometimes engaging in behavior and coverage that comes off as belittling, whether publicly or privately, intentionally or not.

Though it’s often cast in the context of political power—which strategy for engaging the media, both mainstream and new, delivers election victories—this trend of animosity within the sector (because press and creators are a single sector, whether they like it or not) is bigger than that. It’s creating a zero-sum incentive structure that has already led to a degradation of the information infrastructure that is leaving people less informed, and it’s time for everyone to put on their big kid pants in service of the greater good.

A Non-Exhaustive List of Beefs

Attribution

In Twitter’s heyday, when I would publish a new piece, I relished seeing how it was getting shared across the platform. All sorts of folks would distribute and amplify my work: journalists, average users, celebrities. Sometimes I would get tagged, sometimes I wouldn’t. I didn’t care; it wasn’t the reference section to an academic paper. I was just happy to see people learning from my writing.

I think about this when I hear gripes from journalist friends who say that creators are “stealing” their stories, angry that creators are “reporting” news to their audiences they didn’t actually report. I don’t see it that way at all. While there are less-than-scrupulous creators who don’t give credit where it’s due, videos with picture-in-picture or greenscreen screengrabs of the article being referenced are popular and highly-engaged with; often such videos will be a user’s first contact with a particular story.

Similarly, news organizations complain about the lack of editorial standards they see in creator-driven vertical video; as I’ve written before, platform algorithms, which incentivize emotional content, drive some to post sensationalist or conspiratorial videos so they can make rent.

The topline is this: Creators add context and their own interpretation of the news, which helps them connect with their audiences. When creators explain and react to journalistic content, is it “stealing,” or is it sharing? I’d argue it’s the latter—the modern form of a Twitter tag. And if journalists are worried about the lack of referral traffic from these videos or the way the content might be sensationalized, their gripe isn’t with the creators, it’s with social media platforms that disincentivize users clicking links, closing apps and moving off-platform, and reward them for posting rage bait.

Competition

Downstream from attribution is legacy outlets’ and journalists’ concerns that content creators are their competition. This concern is embodied by framing of Reuters’ coverage about the journalist-creator dynamic at the 2024 DNC:

Influencers have access to a creator lounge in the United Center and a creator platform section on the convention floor where they can create video content.

The besieged journalism industry, which has lost tens of thousands of jobs from cost-cutting and consolidation over more than a decade, had its space at the convention slashed compared to previous conventions, according to another Pew study.

“These are the worst working conditions of the 20 conventions I have covered,” said Jonathan D. Salant, assistant managing editor of politics at the Pittsburgh Post-Gazette.

In its 2026 survey of 280 digital media leaders from 51 countries and territories, the Reuters Institute for the Study of Journalism at the University of Oxford found that 70 percent of those surveyed were worried about creators “taking time and attention away” from their work, while only 39 percent were worried about losing talent, in a year when many formerly mainstream reporters made the leap to the land of online creators:

But, according to RISJ’s Digital News Report, which surveys news consumers in 48 global markets, 77 percent of people surveyed consume news online each week, and “most people who get news from creators are using them alongside traditional media – not instead of it. In fact, those who access creators consume more traditional media than the average respondent.”

Trivialization

Creators, for their part, have plenty of complaints about the mainstream media, but by far the one I see echoed the most frequently in their videos is that the press trivializes both the work creators do and its impact. Content creation is treated as a “side hustle” or inherently unserious because it grew on platforms that got their start as a home for viral dance trends. But the platforms aren’t just for silly videos anymore—creators drive behavior change, movements, impact, and even reporting. And, as RISJ’s report shows, audiences find creators easier to understand and more entertaining than traditional media:

I can feel hundreds of pairs of journalists’ eyes rolling—who cares if creators are entertaining?! The news isn’t meant to be entertaining! Sure, we all might like to believe that. But for the past ten years, everyone from Kremlin trolls to MAGA shitposters have been making a bet on infotainment while the rest of us clutch our pearls. We need to work with the information environment that we have, not the one we’d like.

Burying the Hatchet

Earlier this year, serious allegations of sexual abuse tanked Senator Eric Swalwell’s California gubernatorial campaign. Creators Arielle Fodor (“Mrs. Frazzled”) and Cheyenne Hunt first surfaced the accusations; they were then reported out by CNN. As Susie Banikarim of the Columbia Journalism Review writes:

What made the creators’ approach notable wasn’t just their persistence. It was their understanding that social media influence alone would not be enough to take on a powerful politician and care for Swalwell’s accusers. They coordinated with the women who confided in them, helping them find legal counsel and directing them to CNN once they were ready to share their stories.

This example shows that collaboration between the warring camps is possible and productive, and that it can have positive, seismic impacts for democracy.

If I were advising mainstream media about how to compete in this environment, I’d tell them it’s time to bury the hatchet, stop publicly griping about each other1 and start collaborating with content creators and replicating what they do. (The same goes for experts, which I discussed in my recent keynote at GlobalFact.)

Collaboration

For collaboration between legacy media and creators to work—that is, for such work to feel authentic to creators’ audiences—it has to be more than a quid pro quo, “we report the news, you make videos about it and receive payment in return.” This might look like:

1. Reporting a story based on creator tips—like CNN did with the Eric Swalwell story—and involving creators in the process.

2. Inviting creators to shadow journalists for a period of time and make content about the process, pulling back the curtain on the parts of the reporting process that are rarely seen by the public.

3. Hiring creators to skillshare with your newsroom about scripting, shooting, and building and maintaining audience trust.

The uniting factor here is not to treat creators as a sideshow, but as experts in audience cultivation and retention. Recognize what they do is work. Come to them with humility and a desire not to learn from them, not just to extract value from them.

Creators need humility, too. I often see creators telling audiences that “no mainstream media are reporting” whatever their topic is, and nine times out of ten, that’s just not true. Goodwill needs to go both ways.

Empowering Journalists to Make Vertical Video

Remember when posting to Twitter was seen as a necessity for journalists? (For some, it still is—yikes!) Why were journalists never empowered to take the same leap into vertical video? Sophia Smith Galer—who cut her teeth at the BBC and VICE, makes beautiful video explainers for her 1+ million followers, and trains others in how to replicate her success—points out that more newsrooms are prioritizing making vertical video with high production values, but these videos

aren’t coupled with lower production value, faster-to-make videos that are platform native coming from individuals. Content remains glued to publisher accounts across the board and journalists uploading work themselves (like they’ve done on Twitter for years) remains the exception, not the rule.

To me, the gap is structural:

1. Journalists need to know that their newsrooms want them to make this authentic content about their reporting and the behind-the-scenes of it, and they need to know it won’t get them in trouble; this means creating or updating a social media guidebook with basic standards for journalist-generated vertical video.

2. Journalists also need the space and support to make this content. Even when you’re making quick, off-the-cuff videos, it takes time, and journalists shouldn’t be forced to make content in their time off. This should be factored into the workload that newsrooms envision for their journalists, understanding that they might not be filing ten stories a week if they’re also making video, but are doing integral work rebuilding trust with audiences and even reaching new ones.

What’s at Stake

Since 2024, I have been in more conversations than I care to admit in left-of-center circles in which the privileged few—journalists and executives from mainstream media, national security professionals, experts, political consultants—give creators short shrift. We all need to wake up. As much as I’d like to go back to the information environment of the 2010s—ah, those halcyon days!—it’s not going to happen. As wrote in a great essay this week, social media is media now.

My friend Sabra Ayres, who spent 20 years in traditional newsrooms around the world and now trains local journalists in Ukraine—including a course this year for local media on making mobile-first vertical videos—says she’s telling her grantees that “they better get used to creators and try to work with them instead of against them.” I agree. We can each keep trying to push our own boulders of truth uphill amid an avalanche of lies, or we can roll downhill, push the boulder into a lake, and make powerful ripples together. 🧭

Thanks for bearing with me this week! We just got a new puppy—meet Kaylee!—so it’s honestly a miracle I wrote anything at all. I won’t be sending out any field notes this week, but hope to be back to a two-post-per-week schedule when we’re more settled in.

I was surprised to find that my April “Field Notes” post was among the most popular in the six months since I relaunched this newsletter. I’m happy to keep scratching that itch for you!

Every Friday, I’ll send a short roundup of readings to help you find your way through our polluted information environment, along with some “offline” commentary—what’s going on in my garden, my kitchen, my life off screen. I hope you enjoy it...but first, some business!

Reader Survey

Just before we rang in 2026, I rebranded this newsletter. The response has been really positive—there are over a thousand more of you, and I’m thrilled you’re here.

Six months later, I’m still on a mission to make this little corner of the internet as useful as possible for you. Please take this short subscriber survey to help me tailor my work to your needs.

Take the Survey!

Online Notes

A few of my favorite pieces of news and analysis from the past week.

A victory against the “censorship” liars.

My colleagues in the disinformation research space, including and , were granted a dismissal in a case against them filed by a bunch of mendacious jerks and tinfoil cranks (funded by Stephen Miller’s lawfare factory) who alleged they had “censored” them. They had not. The liars found the researchers’ speech inconvenient, and they sued them to try to shut them up. One of the lessons here—which I mentioned in my speech to fact-checkers from around the world last month—is to always challenge the premise of BS suits and investigations, to never roll over and accept failure. The defense team and the folks they represented did just that. They still sustained damage to their reputations and lost funding and in some cases, work. But in the end, the very judge who amplified a lot of the conspiracies about them was forced to eat crow after the Supreme Court rebuked his nonsense, and that feels gooooood.

Celebrating July 4 like I celebrate the Superbowl.

I am feeling a bit melancholy about our nation’s 250th, as I’m sure many Americans are. We’re getting together with some neighbors and their kids and making some food. That will be the extent of our celebrations this year. When did I start celebrating July 4 like I celebrate the Superbowl? That is, when did I start using it as an excuse to gather and eat while ignoring the convening premise? I remember July 4, 2020 being totally surreal; Trump had organized flyovers from a bunch of military aircraft, I suppose meant to distract us from the ongoing pandemic and the ongoing clashes between law enforcement and Black Lives Matter protestors. Maybe that was the turning point?

argues that July 4 feels hollow this year because our leaders don’t respect the virtues, symbols, and values of America’s founding: “They believe that only their clan represents America, that only people like them deserve to be considered “real” Americans, and that the American government exists to serve them alone, they need to undermine anything that tells a more unifying story.”

I hope you enjoy your hot dogs; I’ll be inside and not going anywhere near the travesty on the National Mall.

“I can prompt with the best of them.”

I am teaching my annual summer course on disinformation and influence in the political age. My students are at the Master’s level, and like many other educators, I’m worried about what AI is doing to our collective ability to write, read, analyze, and reason. Rather than wax poetic about that—I’m still planning a longer guide on AI— enjoy this piece of satire on LLMs and the rapid atrophying of our brains.

Offline Notes

The real-life pastimes, products, and programs getting me through the infopocalypse.

It’s Spotted Lantern Fly Squishing Season

While the ideal yard for some Americans is a carpet of green grass, I’ve always wanted something a bit more wild. Growing up, my childhood backyard abutted a few wooded areas where I made forts, and where my dog, Max, would chase deer. (Once, he pulled away from my mom on a walk to chase a herd, only to return hours later dragging a log, which his leash had wrapped around.) We moved in 2025 and have been delighted by all the wildlife in our new area: foxes, skunks, deer, and many birds (our cat, Baxter, is especially grateful for them). We decided to put in some native plants (on which more in a future issue).

After the gardens were planted in May, our big household project has been their care and maintenance. Watering—especially in this heat—is essential, as is weeding and pest control. Spotted lantern flies—an invasive species with no dedicated natural predators in North America—have been feasting on my Sweet Bay Magnolias, so they’ve been the primary pest receiving our redirected rage. The whole family has been enlisted in our assault on them. Every time our four-year-old sees one of the black or red and white nymphs, he shouts, “Mommy! Daddy! A spotted wantern fwy!” We either whack them with a shoe or sandal or flick them into a bucket of soapy water. These scenes of suburban carnage give us great satisfaction—as does the fact that we have seemingly made an impact on the population; where we would once pull a few dozen during each session, those numbers are dwindling.

I scream, you scream, we all scream for sunscreen

Gen Z and Gen Alpha are the subject of a lot of derision for their premature obsession with skincare, but I’m proud of them for doing what millennials and older generations did not: using sunscreen every day. After I gave birth, whatever whackadoodle bullshit happened to my hormones drove some heat-and-sun-induced melasma (freckles on steroids) on my pale-to-translucent skin, and since then I have been pretty obsessive about not leaving the house without a mineral sunscreen on my face and a big, dumb, hat on my head. When I went to Australia, I was impressed with how seriously everyone there takes the sun; young Aussies are required by law to wear hats (they’re part of their school uniforms too, so cute!) and adults will always have one—and some sunscreen—in their bag.

Given that we all now live in the fifth circle of hell, here are my favorite sun protection products that may you protect your skin. (These are all Amazon affiliate links; I’ll earn a small commission if you purchase anything.)

• For gardening, hiking, and time at the pool with my son, I love this extremely dorky wide-brimmed hat with a ponytail hole. Here’s the men’s version. Both are really UPF 80+, lightweight, and dry fast when wet.

• I wear this tinted mineral sunscreen every day; it does double duty as sunscreen, skincare, and very basic makeup. ILIA will match your shade for free on their website. If you don’t like the “dewy” look, this may not be for you.

• If you don’t want a tinted product, I like this face sunscreen mist from Sun Bum.

• I carry this matte, brush-on mineral sunscreen in my purse for touch ups and the rare instances that I’m out without sunscreen on. Pricey, but worth it; I think I’ve had the bottle I’m using now for three or four years.

• Similarly, if you have a kiddo who hates getting sunscreen on, I like this kid’s powdered sunscreen, especially for my son’s face, ears, and the top of his head. It’s also great for babies who can’t use chemical sunscreen yet! One tube of this has lasted us years.

• For the pool and other manmade environments, this is my favorite everyday body sunscreen. It’s not greasy at all. You’ll need a reef-safe sunscreen if you’re snorkeling near any reefs, though.

Widow’s Bay

It takes a lot to make me stay up late these days, but bingeing Widow’s Bay, the comedy-horror show starring Matthew Rhys on AppleTV+ had me doing just that a few weeks ago. There are plenty of gifted TV critics who have written about what makes this show so special, and I’m not them, so suffice it to say it’s been one of the best things I’ve watched in years. It is laugh out loud hilarious while also managing to be legitimately creepy at parts, with excellent writing, evocative set dressing, and physical comedy that is rare in the two-screen era of TV. If you haven’t watched it yet, make time this weekend—you won’t regret it.

Last week, I was honored to deliver a keynote address at the American Association on Intellectual and Developmental Disabilities’ annual conference as the Association celebrated its sesquicentennial (150th anniversary). The topic was personal, and one that is very much in line with the Wayfinder ethos: scams and fraud online, and how this growing scourge intersects with the ongoing war on reality.

Excerpts from the speech, lightly edited for clarity and context, are below.

A little over a year ago, when the AAIDD proposed that I give a speech at your annual conference, I wasn’t quite sure what I would discuss. In some ways, the world my work inhabits felt miles away from yours.

A few months later, those worlds collided.

I got a call from a loved one. She has dyslexia and ADHD and had just gone through a difficult divorce. She said someone had hacked her Facebook account. Could I give her a hand?

She had graduated from high school in the 90s, during the internet’s infancy. Her ex had handled their family admin, and she had worked in customer service for a grocery store for almost 15 years, so she didn’t have regular contact with computers or the internet besides a little bit of Facebook usage and texting friends and family. That meant she had never developed the information literacy necessary to survive in today’s polluted information environment, where disinformation, fraud, and scams thrive.

She gave me her password so I could see what was going on with her account, and I logged in remotely. Even as someone who studies the internet, I wasn’t really prepared for what I saw: she’d been targeted by a man posing as an actor from a soap opera she loved. Over months, he’d cultivated her trust, solicited nude images and account passwords, then threatened her: if she didn’t send him money, the photos would go to her entire friends list. By the time she asked for help, we were fighting in real time to keep control of the account as he blasted those images to her brother and her aunt.

After hours of changing her password only to watch him change it back, I emailed colleagues at Facebook. They were able to help, but I went to bed really upset that night. I had just experienced the two-tiered system of aid at these social media platforms: If you know someone who works there, you have a chance to preserve your dignity. If you don’t, you’re out of luck.

It wasn’t the end of my loved one’s ordeal. Months later, I learned she was withdrawing $20 to $200 several times a week and converting it to bitcoin. She’d met a man on a dating app, allegedly an American serviceman deployed overseas who could only talk in the middle of the night. He’d sent her photos, and she was certain he was real. A reverse image search turned up other women taken in by the same con. The “soldier” had almost certainly used AI to alter his face and voice on those late-night calls. She didn’t know that was possible.

In all, she lost about $40,000 to two romance scams—her entire savings, plus a loan against her 401(k). We reported the incidents to the FBI, to the apps, to her banks. They told us the money was unrecoverable.

It was not her fault. No one ever gave her the tools to see it coming.

Online fraud is growing

Online fraud is a massive industry growing at a frightening pace. The FBI’s 2024 Internet Crime Report logged a record $16.6 billion in reported losses, a 33% jump in a single year. In 2025, that number rose again, to $20.9 billion.

People with intellectual and developmental disabilities are at greater risk of falling victim to online scams and fraud: Australia’s Bureau of Statistics found that people with disabilities experience scams at nearly twice the rate of those without, and identity theft at roughly double. A University of Washington study found 82% of adults with intellectual and developmental disabilities (IDD) named “knowing who to trust online” as a major challenge.

What makes me so frustrated about these statistics is that they are preventable. You do not need to study the internet, or be an expert on disinformation, or have any deep knowledge of these technologies to learn the basics that would have saved my loved one from being scammed.

Government and Platforms: Dereliction of Duty

So why aren’t we teaching these skills? And why aren’t technology companies doing everything in their power to prevent scams and fraud from happening on their platforms? Because words like “media literacy,” “fact-checking,” “information literacy” have been deliberately recast as “censorship.” This is part of a strategy designed by people who have a financial and political interest in keeping you—and the people your community serves—confused, isolated, and vulnerable.

Think about what that actually means. The very skills that would have helped my loved one recognize a romance scam—the ability to verify someone’s identity, to slow down and ask questions, to understand that strangers who profess love and then ask for money are following a script—those skills have been framed as dangerous and antithetical to freedom.

So federal funding for digital literacy programs has been cut. Researchers who study online manipulation have been investigated and subject to frivolous lawsuits. Social media platform safety teams have been gutted. The infrastructure that should have been built to protect people has instead been systematically dismantled, all to the applause of people who call it freedom.

Let me be direct about who benefits from that.

According to internal documents reported by Reuters, Meta projected that roughly 10% of its 2024 revenue—about $16 billion—would come from scam and prohibited-product ads, with high-risk scam ads shown some 15 billion times a day and the company tied to a third of all successful scams in the U.S. That’s a feature, not a bug.

When the companies profiting from scams spend money lobbying against platform regulation, when political operatives attack media literacy as censorship, when federal programs that would teach people to recognize manipulation get defunded, the person who pays the price is your grandmother on Facebook. It is your neighbor who sent his life savings to someone he met online. It is the person with a learning disability who didn’t know the warning signs, because no one ever taught them, and because the systems that should have caught them decided to stand down.

None of this was inevitable. The tools exist. The research exists. The evidence about what works is overwhelming.

What is missing is the political will to use them.

Build the ramp

When I was discussing this speech with the organizers, they reminded me that before the Americans with Disabilities Act, buildings didn’t have ramps. It wasn’t because the technology was complicated. It wasn’t because architects didn’t know that some people used wheelchairs. The ramps didn’t exist because no one was legally required to build them.

As you all know, the burden of access fell entirely on the person who needed it—to find another entrance, to ask for help, to simply not go. Buildings were designed for some people and not others, and that was treated as a fact of life rather than a failure of design.

The ADA changed that. Title III says: if you operate a place of public accommodation—a business, a school, a venue—you cannot discriminate against people with disabilities. You must provide equal access. And if your building has stairs, you gotta build the ramp. Not because it’s charity. Because it’s the law, and because access is a human right.

The internet is a place of public accommodation. But it has stairs everywhere. And we have chosen, as a society, to treat that as a fact of life.

It doesn’t have to be. In 2018, Australia’s eSafety Commissioner began developing Safety by Design, built on a simple premise: the burden of safety should never fall solely on the user. It requires platforms to remove harms before users encounter them, to set safety high by default, to design for people with “distinct characteristics and capabilities,” and to publish real transparency reports about implementation. The EU took a similar path with its Digital Services Act, requiring platforms with 45 million-plus EU users to assess risks to vulnerable populations. The U.S. has nothing comparable.

So here’s the ramp, in three parts. First: a Safety by Design standard for U.S. platforms, with legal liability for the largest ones when they fail. Second: a universal information-literacy curriculum—K–12 and adult—with accessibility built in. Third: targeted resources for people with IDD and the caregivers and educators who are the real first line of defense.

The scam techniques refined on isolated, vulnerable people have become templates for mass manipulation. The information environment left unsafe for the most vulnerable is the same one that radicalized your neighbor.

We are in a war on reality, waged by people who’ve calculated that a confused, isolated public is easier to exploit than an informed one. Every time you teach someone to navigate the internet more safely, you deny that industry a victim. Every time you demand platform accountability, you build the ramp.

Building a ramp is not a political position. Safety is not censorship. Protecting people from fraud is not an attack on free speech. My loved one deserved that support. So does every person you serve. And we can get there—together.

Subscribe now

In a room full of impressive folks whose bona fides include exposing scammers who promote made up miracle cures for autism, holding those perpetrating war crimes and human rights abuses to account, and taking down the biggest purveyor of nonconsensual deepfakes, I argued that the backlash against our work is not a sign of our failure but of our impact. The winning move is to get louder, bolder, and more organized.

Excerpts from my address are below, or you can watch the whole thing courtesy of GlobalFact sponsor LRT, Lithuania’s public broadcaster.

I’ve been asked to speak about what it takes for democracies to win the war on truth—no pressure, right? Just one of the biggest questions of our time.

If I had been asked that question a few years ago, my answer would have been pretty different, and this talk would’ve been a tad more optimistic. Invest in media literacy! Fund fact checkers and researchers and journalism! Regulate unaccountable social media platforms!

Now, don’t get me wrong, those are all crucial components of a winning strategy.

I spent the better part of a decade advocating for them, and if we had seriously invested in them, put the whole weight of government behind them, treated them as the core challenge that they are instead of the “nice-to-have” sideshow some governments pretended they were, we might be in a better situation today.

But it wasn’t solely a lack of investments or a lack of urgency that got us to a point where our work is demonized as “censorship,” where we are facing investigations and lawsuits and funding cuts, where we aren’t just fighting foreign adversaries, but our own allies—my own government!—and those adversaries are engaged in a deliberate campaign to intimidate us and suppress the truth.

That campaign uses the same tactics of suppression that autocrats have always used: media smears, administrative burdens, lawfare, and threats.

To some degree, it was inevitable. When autocrats encounter a challenge to their grip on power and to the narratives that put them there, they will try to eliminate it.

What was not inevitable was our response. We were—and in many cases, remain—too timid, too careful, too technical in a moment that rewards bold action and emotional connection.

Now, I’m not suggesting we fight fire with fire, that we stoop to the level of the autocrats attacking us. We don’t need to traffic in AI slop and online abuse to gain ground in this battle. But we do need to meet people where they are and speak to them in a language they understand—on topics they care about.

We need to take another page from our adversaries, too: we need more solidarity. We need more audacity. We need more courage.

[...]

The broligarchs, as calls them, want nothing more than for us to self-censor, to self-police, to self-select out of the work we are doing in service of the public. In service of democracy. In service of a shared reality.

But my friends, I’m here to tell you, we need to do the opposite.

This is the moment it’s more important than ever to stay in the fight.

If you remember one thing from my speech today, let it be this. We do not need to reinvent ourselves. We did not do anything wrong.

The liars of the world have spent so much time and money attacking us precisely because what we are doing is working. Because we represent a formidable challenge to their power and their profit.

They cut off our access to data because they were worried that our findings would undermine them.

They are using the legal system to keep us preoccupied and scared, to distract us from our work.

They have defunded and decried fact checkers because the facts aren’t on their side.

They are threatening us because we are a threat to them.

So we need to keep fighting. We can’t twist ourselves into knots trying to describe our work in ways they might find acceptable. If we do, we will always be reinventing, rebranding in ways that keep us from connecting with the public we are trying to support.

We can’t bury our heads in the sand and hope to resurface when this is all over. If we do, when we resurface, there might not be anything left to fight for.

The fight is now.

Subscribe now

Here are a few lessons learned from the last four-plus years of my fight:

First, before the worst happens, think through your online and offline security posture. Make sure all the boring details of multifactor authentication and password managers are worked out so that your personal and professional details are safe.

Next, if you become the subject of a politicized investigation—say, from Jim Jordan or one of his copycats around the world—don’t simply comply in hopes that it will go away. Push back. Make them subpoena you. Release your correspondence with them. Demand the transcript be made immediately public or demand a public hearing. Don’t let them operate in the shadows and punish you with process or declare you guilty simply by announcing an investigation. Challenge the premise—and challenge it robustly.

In addition to recommitting to our work, we need to defend it. There are two lessons to learn there.

The first lesson is about strategic, modern, compelling communications. Gone are the days of the ‘spray and pray’ press release. So are the days of considering a mainstream newspaper or TV feature as the pinnacle of influence.

That is what DHS counted on—and what failed it.

If you’re hit with a similar campaign, while you are trying to land the best reporter for your exclusive—or worse, waiting for the storm to blow over and hunkering down, saying nothing—your adversaries will have already filmed and repackaged a thousand media spots and reaction videos and comments. That content will drive more content. You’ll never catch up. Renee DiResta, another researcher who has been targeted by these mendacious lies, likes to say “if you make it trend, you make it true.”

We need to communicate proactively. Particularly in the age of ubiquitous, accessible AI, we can’t wait for a narrative to hit mainstream news or critical mass on social media to get the facts out. Publishing the truth early and often—and priming the LLM answer engines with it, the same way our adversaries are doing—gives us a better chance of undermining falsehoods.

But we can’t just put the truth out and hope for the best. It has to be the truth, well told. When we communicate, we need to eschew the dry and the technical, and communicate like Aunt Sally and Uncle Bob are listening. How would we explain our work to them—and why should they care?

We should also seriously consider if we’re truly the best messenger for our work. To many, we are elite and out of touch. Are there local content creators, community leaders, or independent outlets that might have more credibility with the audience we’re trying to reach? Are we on the right platforms to reach them?

This requires some experimentation. It will lead to failures. But our adversaries have been figuring out what works through trial and error for decades. We should let ourselves do the same.The last, and arguably most important lesson, is about solidarity. We need to focus on the big picture. We need to put aside differences we may have with each other and come together in support of something we all take for granted: our belief in democracy. Now is not the time to debate the perfect definition of disinformation, or to harbor bad blood because five years ago another organization won a grant that you wish you had been awarded. We need to recognize we are all on the same team, and all here to lift each other up.

One of the worst things about my experience has been how isolated I sometimes felt from this community. [...] But we know better than others that if there’s one thing the autocrats of the world are good at, it’s weaponizing division. Let’s not give them any to use against the counter-disinformation and fact checking community. Let’s speak with one voice and defend those who are being attacked. Let’s loudly and clearly and proactively explain who we are and what we are about. And let’s not just say it, let’s show it. Let’s coordinate and cooperate with those who are targeted, even when it might be risky or uncomfortable. We are much stronger together than we are when standing alone.

[...]

So send the tweet.

Use the disfavored word.

Read a banned book.

Publish the paper.

Speak out on behalf of a colleague.

Expose the liars and megalomaniacs profiting from online harms and disinformation for exactly who they are.

Keep on fighting.

EARLY IN the COVID-19 pandemic, I started getting questions from reporters who wanted to know why Russia was claiming biological weapons were being developed at US-funded laboratories in Ukraine.1 It was simple, I told them; Russia couldn’t resist an opportunity to cast doubt on the origins of the coronavirus, especially when it could potentially undermine Western support for Kyiv or make the United States look bad.

Those false and misleading claims, which were peddled as early as 2009, surged after Moscow’s full-scale invasion of Ukraine in 2022, and again in 2024. And last week, outgoing Director of National Intelligence Tulsi Gabbard amplified them again, declassifying documents that she said included “never before seen intelligence revealing new evidence of past US government funding for more than 120 biolabs in over 30 countries, including Ukraine.”

As RFERL reported:

“The unusual move...came just days before [Gabbard’s] departure as DNI, which oversees an intergovernmental office set up to coordinate information sharing among the sprawling US intelligence community. It wasn’t immediately clear why Gabbard was releasing the information. Nor was it clear that it contained anything new or revelatory.”

In a statement that looks like a high schooler wrote it, Gabbard uses classic disinformation tropes, claiming she has blown the lid on a coverup and has been wrongly accused of furthering adversarial narratives:

“The information surrounding the existence, history, locations and funding of these US funded biolabs has been intentionally covered up by powerful people falsely, claiming that they do not exist and accusing anyone who says otherwise to be foreign assets and traitors to America.”2

What the documents actually show is quite different: a mostly-redacted briefing on Russian disinformation regarding a US-funded biolab in Ukraine, and a list of US-funded labs, partners, and funding levels in the country that was far from secret to begin with.

Here are the facts: the United States has openly funded biolabs around the world for decades. The government never denied it. The program, called the Cooperative Threat Reduction program, was never secret—Congress approved it. These labs aren’t developing biological weapons, they exist “to prevent, detect, and prepare for outbreaks caused by biothreat pathogens.” And yes, as the DNI points out, labs in Ukraine “may be at risk of compromise due to the ongoing Russia-Ukraine war” Russia’s ongoing war in Ukraine,3 but that’s because most scientific and defense research is vulnerable to an entirely unscrupulous army undertaking a war of aggression designed to inflict as much suffering as possible, with no regard for downstream effects (see Russia’s destruction of the Kakhovka Dam and its long pattern of threatening nuclear safety and security during the war)!

If Tulsi is worried about this lab, the best thing for her to do would be to work to end Russia’s war. You know, the one President Trump said he’d end on Day One of his second term.

Waving around “receipts” like these documents is in and of itself a time-tested disinformation tactic. Purveyors of bullshit know that most people don’t read past wild headlines, and when they do, they aren’t able to make heads or tails of the underlying source material. (See: the fever dream of falsehoods that is so-called Twitter Files.) They think, “if it’s got a U.S. Government seal on it, it must be trustworthy,” even when the source material says the opposite of what is being asserted.

It’s extraordinarily disturbing that the DNI would engage in this sort of conspiracy baiting and frustrating that Ukraine continues to be a favorite topic for both Russian disinformers and those closer to home. Beyond the continued implications of Tulsi’s lies for Kyiv, I’m most worried about what these actions might mean for the upcoming midterm elections. We’ve seen this administration declassify documents for political gain before: Tulsi declassified documents that she claimed proved the Russia investigation was a “hoax,” along with others she hoped would help the President in his efforts to get his first impeachment “expunged.”

Imagine, for a second, it’s October. What might happen if the next DNI declassifies intelligence assessments about an adversary’s attempted incursions into voting systems? Maybe that adversary is China or Iran, Trump’s favorite bogeymen, which his supporters are primed to hate. Even if the declassified documents only describe ongoing, thwarted interference attempts, or assess with low confidence that an incursion is possible or likely, the administration could wave this intel around to inspire distrust in our democratic infrastructure, or more chillingly, attempt to make changes to the electoral process on the basis of what it says are “urgent threats.”

If confirmed as the new DNI, Jay Clayton—who has been a yes man for Trump and raised questions about the sanctity of elections processes—could continue in Gabbard’s footsteps. There’s not much for ordinary citizens to do about this problem right now, other than be vigilant and seek news and analysis from a range of sources.

One thing is certain: In a world where the nation’s top intelligence official is declassifying documents to win at shitposting, Americans should feel confident in rejecting the lies the Trump Administration tells us—and worried about what that means for our nation’s safety and security. 🧭

Buy My Books

One of the most pervasive myths about online abuse is that it’s “just mean words” that shouldn’t have an effect on us as we move about the world.

My answer to anyone who believes that is to walk a mile in my stilettos and see how you fare.

When your children are threatened, is it “just mean words?” When your likeness is inserted into deepfake sexual abuse material for your neighbors, coworkers, and peers to see, would you be able to act like nothing’s happened? When a stranger approaches you to ask for directions and your entire body tenses in anticipation of what might be coming, did those words stay online?

They’re not. You can’t. They don’t. And anyone who believes otherwise is ignorant to the realities of the pervasive, normalized online harms that women and girls deal with every day.

A new film argues that the harms women face through social media and artificial intelligence are colliding with a backsliding women’s rights environment in an attempt to control and silence half the population. MISS REPRESENTATION: RISE UP, directed by Jennifer Siebel Newsom, is the follow up to the acclaimed MISS REPRESENTATION, which examined how mainstream media systematically distorts and diminishes women’s power over a decade ago. MISS REPRESENTATION: RISE UP turns the same critical lens on social media and AI, wrenchingly encapsulating the reality of being a woman or girl online today.

The film premiered at New York’s Tribeca Film Festival this weekend, and I was thrilled to attend and participate in a talkback panel after the screening. I sat for an interview for the film back in 2024, and was touched that the production team felt my insights and the story of what I went through at the hands of the far right and Fox News were worth including. But the film’s core isn’t archival footage or a parade of experts; it’s the stories of young women making change in their schools, running for office, and crafting policy to protect themselves and others from the harms they’ve experienced.

My mom came to the premiere, and though she’s read my books and watched me go through serious, widespread abuse, the film shocked her. She thinks the film should be shown in high schools across the country. I’m hoping it gets wide distribution on a streaming service. Following the journeys of young women who don’t know a life without the toxicity they face every day on social media will help actualize the offline harm of online violence for the many, many people who still think the problem is thin-skinned women who can’t take a few mean words.

After the talkback panel, which was moderated by MS Now’s Stephanie Ruhle and included Siebel Newsom, former Secretary of State Hillary Rodham Clinton, Dr. Safiya Noble, Jim Steyer, and me, I left feeling empowered. My work is often lonely, but once in a while I get the privilege of standing shoulder to shoulder with people who don’t just SEE what I study, they KNOW it in their bones. They have taken adversity and transmuted it into change. They are dedicated to supporting and platforming each other.

They are the movement builders. The connectors. The mothers.

Amid so much bad news and continued adversity, they are what gives me hope.

—

Siebel Newsom and her team at The Representation Project are building on the lessons in the film to provide resources to those experiencing harassment, exploitation, or abuse online, and to give voice to their stories. You can organize a screening, share your experience, get informed, and get help at missrepresentationriseup.org.

With apologies for being absent from your inbox for a bit, it’s been a really busy few weeks. My son had a birthday—parents know how all consuming birthdays can be for small kids, with the classroom celebration and the party for friends and marking the occasion at home…what a vortex!—I started teaching my regular summer course on “Disinformation and Influence in the Digital Age” at Syracuse University’s Maxwell School, and I’m working toward meeting a very big deadline that’s occupying most of my brain power. I promise I’ll be back to a more regular schedule soon!

In the meantime: do you have questions about AI and how (if at all?) to use it responsibly? I’m preparing a big post on AI, so leave your questions in the comments!

The Future of Truth is Local

A BRITISH FRIEND visited the U.S. recently; when he told a cab driver he lived in London, the man offered his condolences on “what happened to the city.” He believed it was engulfed in crime. The cabbie, it seemed, had been one of millions of social media users to see online disinformation—intentionally false or misleading posts—targeting London and the United Kingdom over the past two years. The users behind these posts tap into people’s real grievances, but not with the intention of solving them; their goal is monetary or political gain.

Research commissioned by London mayor Sir Sadiq Khan shows that since 2024, narratives claiming London is “in decline” have increased at least 150 percent, while migration-related narratives have increased 300 percent. Far from being authentic discourse from the UK about its capital, many users peddling anti-London narratives are unlikely to have ever set foot in the city. The Mayor’s report notes that pro-Russia, pro-China, and pro-MAGA actors have promoted or amplified this divisive content. Last week, a BBC Panorama investigation interviewed the profit-motivated users as far away as Sri Lanka and Vietnam who have shared “anti-immigration AI-generated posts about the UK” to audiences in the millions.

I’ve studied disinformation for most of my career, and even been involved in U.S. government efforts to combat it at the federal level. But the increase in AI-powered falsehoods and our increasingly insular internet have led me to reassess the antidote to this illness. While national-level governments must regulate Big Tech, addressing the economic, societal, and democratic impacts of social media, it is local and city governments, experiencing the first-order effects of disinformation, that should lead the charge to build societal resilience to online manipulation.

The online disinformation playbook has been around for two decades. Moscow used it in Estonia to exacerbate tensions between ethnic Russians and Estonians in 2007; it deployed the same tactics across Central and Eastern Europe for the intervening decade before famously unleashing them in the US ahead of the 2016 election and in the UK during the Brexit referendum. The themes in those campaigns would be familiar to audiences today; in 2016, Russia’s Internet Research Agency purchased ads served to UK and US audiences that contained anti-Muslim commentary from a page called “Stop All Invaders.”

The West’s introduction to this challenge was through the lens of nation-state-level information competition, so its response was nationalised and securitised, focussed on exposing adversarial campaigns that had national security implications. Together with poor communications from governments, this top-down approach has left government campaigns to combat disinformation vulnerable to accusations of censorship, even if such claims have little to do with reality. Further, too little effort and funds have been expended in empowering people to navigate an increasingly complex information environment, a generational investment that would arm them with the skills to identify manipulative tactics no matter what the medium is or which party controls the government.

Throughout my research, I’ve seen that these sorts of programs—grounded in community and aware of local issues—are the most successful. Librarians teaching information literacy. Local not-for-profits showing grandparents how to FaceTime their grandkids and avoid falling for scams. Comedians putting on free shows in their hometowns that both educate about online influence and entertain. Events like these help people understand their information environment. When local officials get involved and neighbors can discuss the issues facing their communities face to face, they also build trust in institutions and strengthen the social fabric.

(If you like this, you’ll love HOW TO LOSE THE INFORMATION WAR! 👇 )

Buy the book

The rise of generative AI makes investing in local approaches to countering disinformation all the more urgent. Large language models and image generators have rendered it trivial for any huckster on the internet to develop bespoke messages and accompanying media tailored to vulnerable audiences. Meanwhile, our once-ubiquitous social media platforms are disintegrating and we’re self-selecting into increasingly insular communities. We need the ground truth that only face-to-face contact can provide.

So at the Cambridge Disinformation Summit in April, when Sadiq Khan highlighted the challenges facing his city as a result of online lies, I was glad to hear a local leader talking about addressing the challenge. He called on tech companies to do their part to stop incentivizing the “division dividend,” and promised his team would be “expos[ing] the people pushing lies about London and debunk[ing] the dangerous lies they’re spreading.” London, he told the audience, would be the canary in the coal mine for other cities around the world targeted by viral falsehoods.

But I think Khan and his city should go further. If he takes the challenge seriously, London could also be the first city with a dedicated resilience-building program, too. The Mayor should evolve his one-off research on London disinformation into a full-time insight and analysis unit that can not only tell him what’s trending, but clue him in before the disinformation reaches a fever pitch as it has done over the past few years. To be most effective, such insight-building should be cross-government, methodically drawing observations from disparate parts of city—and even national—government.

Using that information, his team could not only continue to craft communications campaigns to challenge false and misleading narratives, but understand which audiences in the city are more vulnerable to disinformation. Then, his office could fund community groups to offer trainings for their neighbors on how to avoid being manipulated and scammed online. These groups enjoy more trust than politicians and can mold curriculum so it’s more resonant in the local context. Moreover, the content wouldn’t be political; the same skills that keep you from signing away your life savings to someone you met in a Facebook Group are those that will help you avoid being manipulated by foreign actors. These community events have the added impact of drawing people away from their screens, out of their homes, and face-to-face with their neighbors, with a reminder that there is still kindness and humanity in the world.

Khan, along with New York’s Zohran Mamdani, Minneapolis’s Jacob Frey, and other mayors and governors like them, who are vociferously challenging falsehoods from their bully pulpits are headed down the right path. But they can take it further.

The disinformation spread about these cities is asking residents to “reject the evidence of [their] eyes and ears,” to borrow a quote from Orwell. Winning that battle is more than a he-said-she-said, more than strategically-deployed memes.

It’s about getting people back into their communities—and empowering them to be the first line of defense against online manipulation. 🧭

Paid supporters get access to my brand-new 10-Day Online Safety Tune-Up based on the tactics and tools I used to hold my digital ground against MAGA trolls and Russian hackers. Early reviews called it “invaluable” and “life-saving.” Become a paid supporter and get yours today!

Subscribe now

One of the stranger features of the last four years of my life is that the conspiracy theories about me are so numerous I’ve forgotten some of them. Strangers online purport to know my secrets. They falsely claim I attended Chelsea Clinton’s wedding (no), that I was arrested by a Special Forces team outside of a defunct bar (no), that I was a “clandestine operative in Eastern Europe” (no).

This week, while trawling the State Department’s Freedom of Information Act (FOIA) Archives for a research project, I was reminded of another conspiracy theory about me that burned short and bright in 2024: people believed that while I was getting a resource-poor non-profit up and running and mothering a two-year-old, I was somehow responsible for or involved in the arrest of Telegram CEO Pavel Durov. They believed this so fervently they submitted at least 23 pages of FOIA requests—totaling at least six dozen1— to the State Department seeking my (nonexistent) communications with the Department about Durov or to the French Embassy.

This episode, despite its absurdity, highlights something much more serious: the many ways that democracy’s openness is weaponized to punish enemies and undermine the basic functions of government.

What is the Freedom of Information Act?

Before we get into the meat of this saga, a quick primer on FOIA.

The Freedom of Information Act is a federal law enacted in 1967 and strengthened after the Watergate scandal that gives anyone the right to request access to records held by executive branch agencies of the U.S. government (the courts and Congress are exempted from the Act). Agencies are supposed to disclose requested records unless those records fall into one of nine narrow exemptions for things like national security, law enforcement, trade secrets, or clearly unwarranted invasions of personal privacy. Though all administrations work these exceptions to avoid embarrassing or scandalous disclosures, FOIA continues to be a workhorse of democratic accountability that watchdogs, researchers, and regular citizens use to understand or expose government decisionmaking.2

FOIA is how we learned about the CIA’s post‑9/11 torture program, government surveillance of Black Lives Matter organizers, and emails showing how officials in Flint handled the city’s water crisis.

A Study in Administrative Burdens

FOIA, for all its nobility, is also a system that can be gamed. Increasingly, bad‑faith actors are learning to weaponize the very transparency rules meant to keep the public “in the know.”

Over the past few years, we’ve seen powerful, often well-moneyed interests organizing their supporters to use the FOIA system to impose administrative burdens. Ahead of the 2022 midterms, for example, The Washington Post reported that local election administrators had been inundated with state-level public records requests:

In nearly two dozen states and scores of counties, election officials are fielding what many describe as an unprecedented wave of public records requests in the final weeks of summer, one they say may be intended to hinder their work and weaken an already strained system. The avalanche of sometimes identically worded requests has forced some to dedicate days to the process of responding even as they scurry to finalize polling locations, mail out absentee ballots and prepare for early voting in October, officials said. [...]

The use of mass records requests by the former president’s supporters effectively weaponizes laws aimed at promoting principles of a democratic system — that the government should be transparent and accountable. Public records requests are a key feature of that system, used by regular citizens, journalists and others. In interviews, officials emphasized that they are trying to follow the law and fulfill the requests, but they also believe the system is being abused.

This is an administrative burden, one of a few key tools that autocrats and their supporters use to keep their adversaries down and out. If autocrats can’t sue you, throw you in jail, or frighten you into abandoning your work, they bury you in paperwork, forcing you to invest limited resources in dealing with it. The right-wing ecosystem has used it against the government, but it can also be used against other public institutions.

In the field of disinformation research, publicly-funded institutions like universities have been targeted with frivolous and burdensome requests. I’ve heard over and over again that researchers and public organizations have had to hire extra staff to deal with them. Sometimes, the cost is too simply too great for them to bear. Institutions would sooner shut down entire lines of research inquiry and attempt to insulate themselves from such bad-faith administrative attacks than continue to bleed resources on responding to them.

The Inescapable Gravity of the Conspiracy Vortex

That brings us to the false allegations about my alleged involvement in Pavel Durov’s arrest that gave rise to dozens of State Department FOIA requests.

In the days after Durov’s arrest, a Qanon conspiracy theorist and professional election denier (who I will not identify, as I’d prefer not to drive traffic to their accounts) posted a late night rant on right-wing video platform Rumble. They then cross-posted it to their other social media accounts, including their Telegram channel, which today boasts over 40 thousand followers. The unhinged tirade repeated the core lie about me—that my job in the Biden Administration was to censor my fellow Americans (no)—and took it a few steps further. According to this conspiracy theorist, I was at the “center” of Durov’s arrest in France, and Durov was arrested because he wasn’t “silencing speech” as I allegedly wanted.

(A reminder, for those that may have forgotten: the United States had nothing to do with Pavel Durov’s arrest, which occurred in France and was conducted by French authorities, and the charges against him were not about misinformation, speech, or “censorship.” The French Government indicted Durov “on multiple charges including spreading child abuse images”—you know, that thing Qanon claims to care about!—“[and] drug trafficking” and cited Telegram’s “almost total failure to respond to judicial requests” as it was investigating crimes on the platform.)

The conspiracy theorist then encouraged their followers to send FOIA requests to the State Department that requested my communications about Telegram, Pavel Durov, or with the French Government. This is called “brigading:” coordinated, pre-planned, mass online activity that often gives the guise of grassroots support or dissent about a given topic. The influencer guaranteed their followers that I was behind Durov’s arrest, and that they would be rewarded with troves of my communications if they simply used their rights (and a little help from ChatGPT or Grok) to request them.

And request them they did. Beginning August 25, the conspiracy theorist’s followers sent dozens of similar requests to the State Department. It’s one of the cleanest examples I have that people are willing to take real-world action based on online lies about me.

Needless to say, they didn’t uncover evidence of collusion. They just forced some poor civil servants in the State Department FOIA office to waste their time sending dozens of replies stating no such records existed. They were guaranteed a treasure trove. When they received a goose egg, I doubt they noticed the crack in the mirror of illusion that had captivated them for so long. It’s more likely they interpreted the lack of records as yet another conspiracy, yet more evidence of my alleged crimes. When you’re convinced that everyone is lying to you, the gravity of the conspiracy vortex is practically inescapable.

The saddest part of the story is that though these individuals set out attempting to inflict harm on me, all they did was waste resources: their time, the government’s, and the taxpayer dollars that funded this escapade. This is part of a broader pattern. It wasn’t just local election officials or the State Department slammed with such requests. The Columbia Journalism Review reports that automated systems funded by right-wing money have lead federal FOIA requests to “nearly double” in the past five years. “In that same time period, the number of FOIA officers at the federal level hasn’t budged.”

Defector’s Diana Moskowitz argued:

Faithful mimicry should not to be confused with actual honesty. People who regularly work with public records understand why we have them—because of public pressure on institutions for transparency—and how gleefully those in power (aka state legislators) are about seizing excuses to take that right away. [...] Requesting a large and unwieldy...

...or, say, nonexistent!...

...data set from a local clerk for the 50th time does nothing to address that. What these requestors don’t seem to realize, or perhaps care about, is that they are not only making the lives of election workers much more difficult, but the most likely outcome of their demands is this: Calls for new laws limiting public records requests, which would not benefit the public in the least.

FOIA and other public records laws are critical to citizens’ ability to hold the government to account. I’m glad they exist. (Heck, I’m even glad that my DHS emails have been released under FOIA, because if the requestors actually read them, maybe they would finally realize they were taken in by lies!3 A girl can dream.) I struggle to identify ways public records could even be amended that wouldn’t block or slow down legitimate requests.

That leaves me where I often end up when dealing with issues of truth and trust in modern democracies: building resilience to bullshit. For dozens of strangers to be so convinced, based on one charlatan’s late-night Rumble outburst, that I was involved in another sovereign nation’s arrest of a tech CEO, they had to be pretty far down the rabbit hole. That doesn’t happen overnight. It’s the result of years of feeling disenfranchised, followed by years of powerful or amoral people capitalizing on that disenfranchisement for power and for profit. (The conspiracy theorist who started this saga has multiple paywalled feeds for her subscribers, of course.) We need to help people avoid the mendacious manipulation that is a feature of the 21st century internet—and, when people who safeguard the truth return to government, they’ll need to work hard to earn all Americans’ trust.

It’s not a satisfying or clean answer, I know, but it’s one that endures. As Thomas Jefferson wrote in 1820 (a quotation I’ve cited in many briefings and papers over the last 10 years):

I know no safe depository of the ultimate powers of society but the people themselves; and if we think them not enlightened enough to exercise their control with a wholesome discretion, the remedy is not to take it from them, but to inform their discretion by education. 

Perhaps foolishly, I still agree. It’s part of why this newsletter exists—to help inform your discretion in an increasingly complex, polluted internet. And until we’re all retrofitted with AI-powered chips in our brains, I think it’s a goal worth pursuing. 🧭

This conversation part comms masterclass, part AI panic check, part foreign policy nerd-out, and part theater kid joy session. We start with a discussion of ’s new piece MAGA has effectively captured “soft power” while “Democrats are responding too often with white papers, earnest panels and fundraising emails that sound like hostage notes from a compliance department.”

From there, the conversation zooms into the atomized media ecosystem: hyper‑local pink‑slime outlets like the Michigan “Midwesterner”—which Emily discovered earlier this year— foreign adversaries exploiting the collapse of local news, and audiences that are both more sophisticated and uniquely vulnerable.

It wouldn’t be 2026 without a conversation about AI “slopaganda,” and why the bar for propaganda has fallen so low that openly fake Iranian Lego‑style war videos can rack up hundreds of millions of views and still move audiences. We talk about Anthropic’s unreleased “Mythos” model, the Trump administration’s sudden about‑face on AI safety after ripping up Biden’s AI roadmap, and why betting on “industry will sort it out” is a recipe for both national security risks and more AI sludge in our feeds.

And we talk about the juciest story of the week: Putin’s increasing paranoia about a coup or assassination attempt. We discuss why this story came out now, and the context of United States’ fraying intelligence relationships with allies.

Before we let everyone go, we pivot hard into joy: the Tony nominations. Two former theater kids compare favorite roles and notice a through‑line between what’s being rewarded on Broadway and what audiences are craving online. The shows getting love right now are weird, specific, and gloriously human, not bland IP recycling machines. That feels connected to a broader hunger for work that is obviously made by people, not models.

If you’re feeling lost in our toxic information environment, I hope this conversation gives you both a clearer sense of the big news moments this week, and a reminder that making room for joy is part of how we fight back.

Did you enjoy this conversation? Let us know in the comments and we’ll bring you more national security, tech, and theater commentary! 🧭

Subscribe now

Elsewhere…

• Have you been tuning into The Illegals, the new podcast from The Mayday Network I’ve joined that is rewatching The Americans episode by episode? I’ve been having so much fun diving deep into one of my favorite shows, dunking on Matthew Rhys’s wigs, and discussing the political context that makes this series so perfect. Watch the latest episode here.

• Are you in New York? Next month I’m headed to the Tribeca Film Festival, where I’ll be at the world premiere of Jennifer Siebel Newsom’s new documentary Miss Representation: Rise Up, which examines” the rising backlash against women’s progress and the hostile landscape of technology designed to harass and, ultimately, silence women.” My story is featured in the film. After the June 7 screening, I’ll join a talkback panel that includes Secretary Hillary Rodham Clinton, Dr. Safiya Noble, and Jim Steyers. Tickets can be purchased here.

The first time I was the target of serious online harassment, I didn’t see it at first—I felt it. I was in a doctor’s office waiting room when I noticed my phone getting hot in my pocket. When I pulled it out, my lock screen was filled with Twitter notifications about misogynist comments (RIP phone battery, RIP Twitter).

At the time, my public profile wasn’t huge; my first book had just come out and I had 10,000 or so followers, but I was being targeted by right-wing accounts that had hundreds of thousands. They were saying insane things about me. They claimed I was a psyop (I guess I have that in common with Taylor Swift?), or that I was transgender (a common line of attack against outspoken women). But most of it was just run-of-the-mill women-hating: criticizing my appearance, requesting sexual favors, instructing me to make the unwashed, anonymous, male masses a sandwich. I fought back, which made them angrier.

It was a formative experience for me, driving my interest in quantifying what abuse looked like and qualifying its impact. I ran a study for the Wilson Center and found that over 336,000 pieces of abusive content were targeted at just 13 women running for office during the 2020 election cycle. Women of intersectional backgrounds—who happened to be gay, or Black, or Latina, or another identity—faced more, and worse abuse. I’ll never forget some of the ways our focus group participants described what they had been through. They told us things like: “you don’t feel safe to continue speaking, so you don’t speak.”

Women were self-censoring because men couldn’t restrain themselves from expressing every vile thought they had directly to the women they were about; we actually confirmed that men preferred to yell at their targets rather than just about them. Platforms were failing women, too—their enforcement of their own terms of service was spotty at best and non-existant at worst.

From then on, I became an advocate for targets of online abuse. I didn’t want others to deal with what I had experienced and studied. I wrote another book, How to Be a Woman Online. I taught workshops on online safety for women as far flung as Belarus and Mexico.

I dealt with a second, much worse round of on- and offline abuse that is following me even today. But when anonymous trolls on message boards were trying to hack my devices, when they were threatening me, when they were discussing how to cause me physical harm, a security consultant I worked with told me: if you didn’t have such a good security posture, this would have been much worse.

He was right: I’ve spent most of my career shoring up my digital footprint. I’ve been targeted by Russian hackers, MAGA trolls, cyberstalkers, and creators of deepfake nudes, and traveled and lived in places where state surveillance is the norm.

I had already dedicated a lot of my career to helping women hold their digital ground, but having put the skills I’ve developed through the ringer, I knew at that moment I would always take great joy in sharing them with others.

Unfortunately, I encounter people in every corner of the world who think that basic cybersecurity precautions “aren’t for them.” They think because they’re just a normal person, with a normal job, hackers and scammers aren’t trying to access their information, steal their identities, or worse—discourage them from participating in public life. That’s not true.

It’s also not true that you need a certification in cybersecurity to be measurably safer online. That’s where the 10-Day Online Safety Tune-Up comes in. Over the next ten days, you will take concrete steps to deepen your digital defenses, so that if the worst happens, you won’t be left wondering what to do. You’ll be safe in the knowledge that you took proactive steps to blunt the impact of harmful digital campaigns with The Wayfinder. 🧭

Read more

Subscribe now

The first time I learned about the Internet Archive and its Wayback Machine—a library of over one trillion snapshots of webpages since the earliest days of the net—was in my high school AP US Government class. It was the early 2000s, and Mr. Fenster, our teacher, was showing us how to access a slice of history. Though I’ve long forgotten the exact website we were looking at—an early version of nytimes.com, maybe?—I remember being bemused at how the internet used to look.

Today, I rely on the Internet Archive for my research. I have a Wayback Machine sticker on my laptop. The Archive has helped me understand the impact malicious lies might have on my safety and allowed me to track changing narratives from both state-backed actors and independent grifters. It has tried to equalize access to information.

But today, the Internet Archive’s—and by extension, the world’s—access to 23 major news sites is under threat. It’s not because they are worried about paywall hopping. Their decision is motivated by something much dumber: the AI industry’s insatiable appetite for information.

Buckle up, friends—generative AI isn’t just hallucinating “facts” and creating deep fakes anymore. Now, it’s directly undermining the historical record.

What is the Internet Archive?

The Internet Archive is best known for the Wayback Machine, which allows anyone to take a snapshot of what a webpage looked like at a given time, preserving that moment for the future. If a news outlet issues a correction or a politician makes a surreptitious edit to their website, the Archive provides a definitive record of a website at a specific moment, so long as someone—or an automated crawler—archives it.

The Archive’s founder, Brewster Kahle, says his goal in founding the Archive in 1996 was “to be a record of what happened so that people can’t rewrite history.” That goal has become a pressing need under the second Trump Administration, which has worked to suppress inconvenient facts and censor information it doesn’t like.

Less well known are the Archive’s libraries of 2 million books and 3 million hours of television, including an archive of state-run propaganda from the likes of Russia...and corporate propaganda from the likes of Fox News. The latter helped me understand the extent to which Fox had lied about me in 2022-2023. Since I don’t watch much TV, let alone keep Fox on all day and night, I needed a way to understand how often Fox had talked about me. The Archive provided it. By my accounting, I was among the top 20 Democrats that Fox discussed, even months after I left government service.

That’s a personal example, but the professional list is long and far-reaching. I’ve used the Archive to access content that has been taken down. Kat Tenbarge writes, “I’ve used the Wayback Machine to visit archived webpages from more than a decade ago, some of which helped corroborate serious allegations. I use the Wayback Machine regularly to figure out how many followers someone had at a given time, to resurface since-deleted material, and to find context for historical posts.”

In my research, I regularly Archive public interest links myself in order to preserve access to them and create a historical record. I’ve been burned too many times by dead links in other people’s work, and that’s not their fault; Kahle points out that the life of the average webpage is only 100 days.

Why Are News Sites Blocking the Wayback Machine’s Crawlers?

According to reporting from WIRED, it’s not the Internet Archive itself that has led 23 news websites to block the Archive’s crawlers. It’s AI:

USA Today Co. spokesperson Lark-Marie Anton emphasized that “this effort is not about specifically blocking the Internet Archive” but instead part of the company’s broader efforts to block all scraping bots. Robert Hahn, the Guardian’s director of business affairs and licensing, says that it has been in conversation with the Archive over “concerns over potential misuse by AI companies of content sets crawled for preservation purposes.”

In plain English, there are two worries at play here:

1. AI models need new data to improve, and they’re using web crawlers not dissimilar to the Archive’s to hoover up that data. The Archive’s web crawler is collateral damage.

2. News organizations (rightfully) want to be compensated if their work is used to train AI models. Instead of (or sometimes, in addition to) concluding agreements with those organizations, AI companies may be scraping the Archive’s data. News organizations are (again, rightfully) angry.

While the Internet Archive recently told WIRED it is “in conversation” with the New York Times and other outlets about a resolution, those outlets have been issuing mealymouthed comments to the press that appear to be the PR equivalent of “it’s not you, it’s me.” The Archive isn’t the problem, they’ve said, it’s those bad AI companies! Now, I’m not the most technical person, but I imagine it would be relatively easy for news organizations to whitelist the Archive’s crawlers while blocking the malicious ones from scraping their sites. Instead, they’re choosing the nuclear option at a time when the Archive’s work couldn’t be more important. These organizations are also shooting themselves in the foot. Even as USA Today Co announced it would be blocking Archive crawlers, its journalists relied on the service for a recent investigation on ICE immigration data.

As journalists organizing to support the Archive wrote:

In previous generations, journalists would turn to the physical archives of a local newspaper or of a local public library to access historical reporting and follow the threads of the present back into history. With many newspapers closed, and no clear path for local public libraries to preserve digital-only reporting, the work of safeguarding journalism’s record increasingly falls to the Internet Archive.

What You Can Do

I assume you are as incensed about these attacks on the Internet Archive as you would be if you learned your local library couldn’t acquire books anymore. That’s essentially what’s happening here—along with the undermining of current and future journalism, accountability, the historical record, and the truth.

But you can push back:

1. Archive stuff. I’m going to require that my graduate students archive sources for their papers, and as a rule, I do my best to archive sources I use in my work or shady stuff I see online. You can do the same. Simply add the url to the box on the right hand side of web.archive.org.

2. Donate. The Internet Archive is a non-profit. They’re being attacked by extremely powerful forces. Need I say more?

What’s your earliest memory of the Wayback Machine? What do you use the Internet Archive for? Let’s make the comments a love fest for an OG internet gem. 🧭

While at the Summit, I ran an online safety workshop. One participant called it “truly invaluable” and another said the content I presented was “saving lives.”

You can have access to this life-saving advice, too! Paid supporters of The Wayfinder will receive my 10-Day Online Safety Tuneup—a toolkit based on my Cambridge workshop—as a perk before the month is out. (All my posts remain free for everyone to access; this is a thank you to those who make that possible.)

If you haven’t become a paid supporter yet, now’s a great time to do so. The Online Safety Tuneup will help you do your digital spring cleaning a little bit each day, but I promise it will be much more pleasant than cleaning your patio furniture.

Subscribe now

Is Iran winning the propaganda war? — a new piece in The Nerve

After the Summit, I got to meet with the inimitable in London. I’ve admired Carole—who broke the Cambridge Analytica scandal, who has long reported on the “right-wing fake news ecosystem,” who coined the term “Broligarchy”—forever.

Last September, Cadwalladr and four colleagues formerly of the Observer (for Americans: the world’s oldest Sunday newspaper) founded The Nerve after the Observer was sold to Tortoise Media. (Read her account of the end of that journalistic era.) They launched the site—which covers politics, culture, and tech—with their redundancy payouts.

Sarah Donaldson, who leads the publication, told the Press Gazette that “the name The Nerve had been chosen because:

we realised nerve is the essential quality needed in our increasingly turbulent world. Too many people in positions of power are losing nerve, and the people we most admire have it in spades. We want to channel that kind of courage into a new publication. As our lives become more and more dominated by AI and algorithms we love that the name feels human and captures the idea of connecting people.

Amen. Nerve has certainly fueled the last four years of my life, so I was honored when Carole asked me to write about those crazy AI-generated Iranian Lego propaganda videos. Is Iran winning the propaganda war? I think so.

They’re crass, they’re childish, and they’re beating Trump at his own game. Iran is winning the information war, and MAGA only has itself to thank...The Trump administration has governed by and for trolls. Now Trump and company are discovering that their tactics are replicable on the increasingly lawless internet they’ve championed, particularly when they resonate with a pissed-off public.

Read the rest of the piece here, give The Nerve a follow on your social media of choice, and sign up for their twice-weekly newsletter. Their content, like mine, is free to access, so support it if you find it valuable.

Finally, send me your questions for another edition of Ask Nina, your digital advice column! Do you have an aunt or uncle who has fallen into an online vortex? Have an online safety question? I can help you (or a ‘friend’ 😉) find their way through this complex information environment. Don’t be shy! Submit here.