President Trump is about to appoint an "AI czar" to run a new "AI Force" — and whoever gets the job will decide what the technology is allowed to do and, critically, what it's allowed to say.

The president announced Sept. 19 that he would name a new AI adviser and stand up the AI Force, putting one person at the controls of policy governing a trillion-dollar industry. That means gatekeeping power over the models millions of Americans rely on for information, with OpenAI and Anthropic already lobbying for more government intervention and lawmakers pushing to regulate frontier AI labs. The question isn't whether Washington will regulate AI — it's who writes the rules and whose interests they serve.

Venture capitalist David Sacks held the AI and crypto czar post briefly before hitting the 130-day special government employee limit in March. His departure opened the door to a broader power struggle. The New York Post reported four names in the running:

Scott Bessent. The Treasury Secretary is the reported frontrunner and has been pushing AI companies to bear liability for what their models produce. "The best way to guarantee safety is that the creators are liable for what they build and generate," Bessent told the House Financial Services Committee. He's also been negotiating with Chinese officials on a mechanism for communicating about AI incidents that threaten national security. Bessent called a Reuters report that Trump already asked him to serve "fake news." A Treasury Secretary who wants AI creators liable for outputs is a man positioning to control speech infrastructure — liability regimes are how governments force platforms to censor without passing a law.

Sriram Krishnan. The senior White House AI policy advisor was named one of TIME's "Architects of AI" in 2025. A venture capitalist from Chennai, India, with stints at Microsoft, Meta, and xAI, Krishnan designed the administration's AI Action Plan — a blueprint to roll back regulation and promote data center construction. Trump praised him at a holiday party, saying "without him things, certainly on AI, would not function well." Krishnan took on a formal NEC role after Sacks left. He's the deregulation candidate.

Sean Cairncross. The National Cyber Director has been embedded in AI policymaking since 2025. A former Republican National Committee COO and Trump-appointed CEO of the Millennium Challenge Corporation, Cairncross has clashed with Bessent over Treasury's involvement in AI matters. One anonymous official told Politico that Cairncross was "in over his head" regulating frontier labs. A party operative running cyber and AI policy is the revolving door in plain sight.

Michael Kratsios. The White House OSTP director has argued against slowing frontier AI development. "The frontier of intelligence is advancing rapidly. That is not a reason to pause its further development or to constrain it with new global governance structures," he said.

The fault line is clear: Bessent wants liability — which means control. Krishnan and Kratsios want to let the market run. Cairncross wants the portfolio but may lack the chops. The Post framed the race as a "power struggle"; it buried the free-speech stakes entirely. What none of the candidates are talking about is what happens when the czar decides an AI model's output is "unsafe" — and who gets to define that word.