A Pro-AI Super PAC's Secret Meme Sockpuppets

Written in collaboration with User Mag.

Last July, ten days after Build American AI, a dark money group tied to pro-AI super PAC Leading the Future, filed its articles of incorporation in Nevada, a new meme page called “Doomers are dumb” cropped up on X. At first, it appeared to be a standard meme account attempting to go viral by posting generic content like "Gay son or thot daughter?" and joking about getting cheated on by their waifu. But in October, around the time Build American AI began publicly campaigning, the account pivoted. Suddenly, it began posting relentlessly about AI and sparring with “Jonathan Doomer,” a self-proclaimed AI skeptic who says he quit his job to spread warnings about the technology on Twitter (and who also began posting about AI on the exact same day).

Despite their differences, both the pro-AI meme account (@DoomersAreDumb) and purported anti-AI activist (@JonathanDoomer) seem to have one thing in common: they both love promoting Build American AI. @DoomersAreDumb directly responded to a Build American AI tweet from the day it publicly launched, and retweeted Build American AI's launch recap. @JonathanDoomer was also in @BuildAmericanAI's replies. To date, each of these posts from the dark money group has fewer than 500 views.

With over $125 million raised and over $70 million in cash on hand, Leading the Future, and its affiliate Build American AI, have become some of the AI industry's most powerful political organizations. Leading the Future's two largest donors are OpenAI affiliates: Greg and Anna Brockman, the President of OpenAI and his wife, gave $25 million to the PAC. Another $25 million was given by Marc Andreessen and Ben Horowitz, co-founders of the venture capital firm a16z, which has invested in OpenAI and co-led its most recent $122 billion funding round. Leading the Future and Build American AI are working overtime to shape elections, influence AI regulation, and build support for an aggressively pro-AI vision of the future. 

Their methods, however, have come under scrutiny. In April, a Leading the Future vendor was revealed to have been operating fake news sites run by AI journalists to attack their critics. In May, a WIRED investigation revealed that Build American AI was secretly paying influencers to spread pro-AI messaging and stoke fears about China. Another anonymous X account with links to Leading the Future spent months attacking OpenAI’s critics with misleading claims and paid ads. And, more recently, posts by multiple Leading the Future affiliates have been mysteriously boosted by accounts belonging to OnlyFans models. 

As these tactics have come to light, OpenAI has sought to distance itself from Leading the Future and Build American AI. "We want to be explicit: No outside political group speaks for OpenAI or represents our company’s views," OpenAI published in a blog post on Monday. "Groups that are advocating on AI should be clear about their policy views, be honest about whom they represent, and not use tactics like astroturfing that obscure the real choices facing policymakers and the public."

Though “Doomers are dumb" and Jonathan Doomer have both struggled to get traction, amassing less than 2,000 followers total between them, Jason Kwon, OpenAI's chief strategy officer, follows them both. So does Josh Vlasto, Leading the Future’s top operative. A handful of other individuals linked to Leading the Future follow at least one of the accounts, and both of the sockpuppet accounts have engaged heavily with Build American AI and its affiliates, including Nathan Leamer (executive director of Build American AI) and Collin McCune (Andreessen Horowitz’s head of government affairs).

A Pro-AI Doomer 

Though Jonathan Doomer posts as a staunch anti-AI activist now, the account's early posts reveal that the page wasn't always opposed to AI technology.  On October 22nd, when both Jonathan Doomer and “Doomers are dumb” pivoted from generic memes to posting about AI, they were both spreading pro-AI messages aligned with Leading the Future's interests. It wasn’t until November that the account now known as @JonathanDoomer changed its handle and began posting negative tweets about AI.

Since then, the page has claimed to be run by an engineer with a 20-year career who became worried about AI and has since quit his job to warn about its dangers. It has posted over 700 times, in some cases generating hundreds of thousands of impressions. Jonathan Doomer generally espouses extreme opposition to AI technology, even though Pangram classifies the majority of his longer posts as AI-generated.

It is likely that Jonathan Doomer is not a person at all, but yet another alter ego of a man named Jason Levin, whose company Memelord Technologies appears to be running the Leading the Future-affiliated sockpuppet meme accounts. Jason Levin gained fame last year from working briefly as Andrew Cuomo's "right-wing memelord" in his mayoral runoff against Zohran Mamdani. Now, Levin does digital marketing and operates a platform that provides trending meme templates and AI content creation tools. Josh Vlasto, Andrew Cuomo's former chief of staff, is in charge of Leading the Future's messaging efforts. 

Much of the early engagement on “Doomers are dumb” and Jonathan Doomer came from other Memelord-affiliated Twitter users. Memelord founder Jason Levin was interacting with both accounts on the first day they began posting and has continued to do so substantially since then. Other memelord-affiliated accounts have played a similar role in manufacturing engagement for the new Doomer-focused pages, including @ChatgptLunatics, @LinkedInLunat1c, and @techbromemes. 

@JonathanDoomer often deploys the same sort of inflammatory language that has been blamed for the attacks on Sam Altman by various figures associated with the super PAC. One post from the account, made two weeks before the attack on Altman, depicts an assault rifle and the phrase “we don’t call 911” in response to a warning about AI. Another post compares Sam Altman to a caterpillar that “murders insects” and “stitches their dismembered body parts onto its back as armor.” Various other posts from the account call supporters of AI “assimilated” and “organic drone[s],” claim that we are less than 12 months away from an AI-induced apocalypse, and boast about mailing dryer lint to the corporate headquarters of tech companies in an attempt to sabotage data center HVAC systems.

Jonathan Doomer has also supported the super PAC’s worldview by adopting the straw-man positions it accuses its critics of holding. Various posts adopt a pro-China and anti-U.S. position, including one that reads: “They say if we slow down, China will win the AI arms race. I say let them. China has much better quality of life than the U.S. China should be the world leader.” Similarly, when New York’s governor Kathy Hochul signed the RAISE Act into law (a bill that Leading the Future once attacked but now claims to have supported), the false flag doomer account tweeted in ironic support of it, saying “People are mad the RAISE Act might slow innovation. Correct. That’s the feature.”

Cannibalism, Mossad, and The Perfect Foil

In addition to discrediting AI skeptics by adopting conspiratorial or strawman arguments, the account serves as a perfect foil for the pro-AI @DoomersAreDumb account to roast. @DoomersAreDumb has engaged with Jonathan Doomer multiple times, usually responding to it with quips like “ngl you’re pretty dumb” or another that reads “it appears the educational system failed you. Maybe an AI tutor would have helped you develop properly.”

The anti-AI sockpuppet isn’t the only “doomer” that Memelord’s pro-AI account roasts. It also goes after a cast of real people and organizations. One post uses videos of real people with Down syndrome to attack the protest group Stop AI. Other frequent targets include Eliezer Yudkowsky, Roman Yampolskiy, and Bernie Sanders. Two of the most revealing targets are Alex Bores (the politician that Leading the Future is spending millions to defeat) and Dario Amodei (the head of Anthropic, OpenAI’s biggest rival, and a supporter of a super PAC opposing Leading the Future).

The posts are often in poor taste. The account has made posts mocking disabled people, disparaging the appearance of real women, and generally containing lewd or offensive content, such as Kermit the Frog masturbating or references to deepfakes targeting intimate partners. Other posts from the account casually joke about swallowing semen, cannibalism, and Mossad. 

If Leading the Future is paying for this, Memelord likely would not appear as a vendor in its FEC filings. Such an expenditure could easily be filtered through dark money channels such as Build American AI which, as a 501(c)(4), doesn’t need to disclose spending. Leading the Future has also been accused of leveraging shell vendors to obfuscate its spending. The links between the meme accounts and the super PAC and dark money group are too overwhelming to be happenstance, however Leading the Future declined to clarify the nature of its relationship to the accounts. OpenAI also declined to comment.

No Grey Area

The use of such offensive and violent messaging, as well as the creation of a bona fide false flag activist, is atypical for even the more cynical political operations in DC. "There's not a grey area here, without clear disclosures about the funding and whose messaging is being promoted, the public is left in the dark," Jamie Cohen, associate professor of media studies at Queens College, CUNY said. "These messages become mainstream." 

AI technology is already radically upending our information ecosystem. The public is swimming in a sea of false information and undisclosed marketing. Politicians, corporate lobbyists, and bad actors are running 24/7 influence campaigns while influencers and meme account administrators are selling off access to their platforms to the highest bidder. 

All of this is incredibly corrosive to democracy. While Leading the Future attempts to appear professional to its contacts on the Hill, promoting a positive vision of a democratic and pro-innovation future where AI tech benefits everyday people, its ties to anonymous influence campaigns, deceptive online personas, and offensive meme accounts tell a different story. 

In light of these controversies, it’s understandable why OpenAI would want to distance itself from the super PAC. But it is hard to reconcile the deep connections between the two entities. Leading the Future is primarily funded by OpenAI’s president, Greg Brockman, alongside his wife and the lead investor in their most recent funding round. Greg Brockman has said that this donation was made in service of OpenAI’s mission. It’s also been revealed that Brockman’s political donations were advised by Chris Lehane, OpenAI’s chief lobbyist, who played a role in the creation of the PAC (which is itself modeled on a PAC Lehane previously helped create). 

The arrangement seems designed to let OpenAI have its cake and eat it too: reaping the benefits of political spending in DC, where many lawmakers will see the PAC as synonymous with the company, while also creating distance to insulate the company from the reputational costs that come with heavy spending on bare-knuckle tactics in the pursuit of a light-touch regulatory environment.

If the tech billionaires and AI-boosters funding and operating Leading the Future and Build American AI truly believe that artificial intelligence will transform society for the better, they should be willing to make that case openly and transparently. Whether Leading the Future financed these meme accounts or merely cultivated the network that produced them, the result is a polluted information ecosystem that warps the public perception of AI. As lawmakers debate the future of this technology we all have the right to know exactly who is shaping that conversation and whose interests they are ultimately serving.