Cathie Wood strongly supports David Sacks’ view: Reports claiming “AI will destroy humanity” are orchestrated.

Cathie Wood noted that David Sacks, co-chair of the U.S. President’s Council of Advisors on Science and Technology (PCAST), believes recent reports about "AI destroying humanity" may be orchestrated. The comments originated from a long-inactive X account belonging to an Anthropic employee who left the company just six weeks after joining, and the content quickly made its way into mainstream media. Sacks also named the parties involved in the incident and criticized The Wall Street Journal for its role. Recently, Anthropic CEO Dario Amodei published a lengthy essay calling for slowing the advancement of frontier AI capabilities, arguing that AI is increasingly involved in developing the next generation of AI. Paired with multiple recent incidents of AI agent malfunctions, he added, security research is failing to keep pace with capability growth. Sam Altman soon publicly backed the call, stating that OpenAI has also been discussing "pacing the frontier" in recent weeks, and announced the company will introduce similar long-term independent assessors. Sacks, former White House AI and crypto chief and current PCAST co-chair, then fired back: "If you truly believe internal models are dangerous, shut them down yourselves." He argued that requiring government approval for the largest AI firms to collaborate on setting speed limits would amount to a cartel, and also singled out METR, questioning whether it is fit to serve as an independent referee for the entire industry. METR has long conducted external evaluations for companies including Anthropic and OpenAI, and recently hired Joe Benton, an Anthropic security researcher who left the firm. However, METR clarified that it does not accept funding from these frontier AI companies, only model access and free tokens.