a16z Partner Casado: AI Opinion Warfare Funding is Unclear
a16z General Partner Martin Casado shared a post about X accounts, stating it makes him appear "darker and more powerful," and mentioned he cannot see where the money is flowing but feels that Brian Chau owes him back pay.
The shared content comes from Tyler Johnston: In recent weeks, a group of X accounts has repeatedly claimed that those expressing concerns about artificial intelligence are merely pawns, backed by well-funded psychological warfare orchestrated by powerful interests. Casado did not name the accounts but humorously directed his criticism at Chau, who has long tracked the "doomsday narrative funding chain."
Johnston's Midas Project has repeatedly written about the connections between OpenAI and a16z's affiliated super PAC, Leading the Future, and its subsidiary, Build American AI: anonymous parody accounts, paid promotions, and outsourced meme accounts. The super PAC has reportedly raised over $125 million. Build American AI has admitted to hiring external vendors to operate parody accounts, stating that it is not a core strategy.
On the other side, Chau has pointed to funding for security narratives from Coefficient Giving (formerly Open Philanthropy), the Tarbell Journalism Program, and the Future of Life Institute, noting overlaps with early investors in Anthropic. a16z, where Casado is a partner, is one of the major funders of Leading the Future. He has recently discussed the new risks of scale laws and centralized computing but has long opposed rapid singularity-style doomsday models.
This is not about regulatory filings or funding flow disclosures, but mutual accusations on the same platform: one side claims that the security faction is funded by foundations, while the other says the accelerationists use dark money and sockpuppet accounts to attack opponents. Both sides treat "who pays" as the primary variable for explaining public opinion.
In market mechanisms, the buyers are laboratories and venture capitalists regulated by federal and state legislation, while the sellers are intermediaries who can turn narratives into posts, memes, and journalist salaries. Event drivers include data center site selection, state regulatory preemption, and the upcoming midterm elections. Funds flow into super PACs, dark money organizations, and content outsourcing; the beneficiaries are those who can simultaneously hire accounts and journalists, while those under pressure are researchers without equivalent budgets who can only rely on public papers for speaking out.
Source: Public Information
ABAB AI Insight
Casado's remark about "back pay" intertwines two funding chains into a single joke. The a16z and OpenAI camps buy political ads and grassroots appearances through Leading the Future; the security side pays journalists, evaluators, and advocacy organizations through the Open Philanthropy system. Neither side is posting spontaneously; both treat the policy market as a purchasable attention commodity.
Capital paths thus split into two accounts. One is the super PAC plus vendors, aimed at suppressing state-level regulation and portraying opponents as part of a Chinese narrative or doomsday cult; the other is effective altruism foundations plus scholarship journalists, aimed at embedding risk assessments into mainstream media and timelines. Chau specializes in the second account, while Johnston focuses on the first. Casado stands within the funding institution of the first account but uses the accusatory tone of the second, effectively acknowledging that "who pays has the right to explain the other" has become industry practice.
Analogous examples include energy lobbying against environmental foundations in climate debates and super PACs against consumer protection organizations in the crypto cycle. The industry phase is characterized by the spillover of model capability competition into legislative preemption battles. a16z itself admits that its policy team amplifies specific accounts, while OpenAI attempts to separate personal donations from corporate positions, indicating that this layer of operation has become too costly to deny.
Structural judgment pertains to the transfer of pricing power. The social pricing of AI risks is no longer primarily determined by papers and incident records but by who can continuously purchase account networks and journalist positions. The mechanism is: when both sides portray opponents as hired pawns, real empirical evidence is downgraded, and budget size becomes a matter of interpretive authority; money does not need to prove a viewpoint true, only to make the other side appear unclean.
ABAB News · Cognitive Laws
First ask who pays, then ask if the viewpoint is true.
When both sides have dark accounts, pawn accusations will counter each other.
In the policy market, meme accounts and journalist scholarships are the same commodity.