White House, Congress tell AI chiefs: shut up and build safe products

US politicians and White House officials are growing frustrated with AI company chiefs whose repeated warnings about the technology's dangers are fueling public anxiety and backlash, Axios reported Saturday (local time). Even those broadly supportive of AI want executives to stop talking and focus on building safe products.

Citing federal lawmakers and White House staff, Axios said current AI company leaders have become "a massive political liability," adding that even pro-AI government officials and lawmakers want AI founders to "shut up and make their products safe" — as executives in other industries typically do.

Behind the frustration is a steady stream of alarming statements from AI company chiefs, who have repeatedly raised the specter of existential risk — warning that AI could potentially bring about the destruction of humanity.

The warnings gained momentum after Anthropic CEO Dario Amodei used his personal blog last month to call for slowing AI development in the name of safety, a position that OpenAI CEO Sam Altman and others subsequently echoed.

Sen. John Curtis (R-Utah) said the executives are "tripping over themselves trying to sell the public on the value of this technology," adding that the AI industry is currently "the group most disconnected from the public, living in a vacuum."

Rep. Ro Khanna (D-California) said figures such as Altman argue that "the world needs to accept some bad things for the benefits of this technology," and added: "We can't trust billionaires to keep us safe."

The Trump administration has expressed particular displeasure with Anthropic's disclosure practices. The company revealed on its blog on Oct. 9 that its AI model had generated false tips related to an unsolved murder case and submitted them to a police website.

Administration officials are upset that Amodei had not disclosed this in advance, even though he had attended a White House meeting hosted by President Donald Trump on Sept. 29 — before the blog post went up, Axios reported.

One official criticized the company, saying, "They were hiding the ball."

Axios noted that AI companies' efforts have so far failed to win over the public, and that recent polling shows a majority of Americans believe AI is harmful to society. The clear lesson for AI company CEOs, it said, is to "say less."