Anthropic and OpenAI hire at least five former Trump officials as AI firms seek favor with the White House

Michelle Nie Duncan, a visiting fellow at the Center for a New American Security, described the exposure to CNBC. "Frontier AI companies are aware that the Trump administration has a lot of levers over their business, from federal procurement and supply-chain designations to export controls and everything in between, and that it's willing to use them," she said ^[8]. Trump has called AI safety concerns a "hoax". He has also said the Department of Justice and other law enforcement bodies could "rein" in AI ^[16]. "They have learned from the Anthropic episode that it pays to be on the good side of the White House," Nie Duncan said ^[9].

Huang's background shows where Anthropic expects federal policy to touch its business. His website says he was responsible for leading the "AI team at OSTP & coordinating national security, science policy, and semiconductor issues" ^[4]. At Anthropic he will work on infrastructure expansion and "planning for rapid AI progress," he said on X ^[5]. The company struck a roughly $45 billion cloud deal with Nscale, CNBC reported in August ^[7]. Anthropic has now put a former White House semiconductor coordinator inside a spending plan measured in tens of billions of dollars.

One reading is a hedge on valuation. Two labs heading toward what Nie Duncan called "some of the largest IPOs of all time" ^[11] are paying for protection against a lever that has already been used on one of them. Her interview also supports a second reading. She said the labs can often offer compensation beyond what is possible in the public sector ^[12], and that many officials "also see this as a chance to be part of something historic" ^[11]. On that view the officials are chasing equity, and the hires tell you more about the job market for policy staff than about how investors will price federal risk. A third reading is that the national security work is simply the job. "[Lind's] expertise is even more important as AI becomes central to the national security agenda, with the Office of the National Cyber Director playing a growing role in shaping AI policy," said Sasha Baker, who leads OpenAI's national security policy team and was deputy under secretary of defense for policy in the Biden Administration ^[14].

I think the hedge reading fits Anthropic and is weaker at OpenAI. Anthropic is the lab carrying the label the court upheld, and its chief executive, Dario Amodei, joined a lunch between Trump and AI bosses and dined with Trump in September ^[17]. OpenAI's hires work across strategy and special projects ^[2]. Lind's cyber brief fits the national security reading at least as well as the access one. Counted together, the two labs have taken on at least five former Trump officials ^[19].

CNBC's report does not put a dollar figure on what the supply-chain label has cost Anthropic. Suppose the labs' eventual IPO filings list procurement and export controls as routine risk factors and tie no revenue to a federal decision. Then the hires were mostly about careers and policy work, and the valuation thesis fails. OpenAI directed CNBC to earlier statements about two of its hires, and Anthropic and the White House did not respond to requests for comment ^[18].