At the last minute, Tear Up AI Safety Order and Bring in Former Attorney General, Trump Vows to Beat China in AI Race with Zero Regulation

According to Dynamic Beating monitoring, the United States' artificial intelligence regulatory policy has taken a sharp turn. President Trump, just hours before the signing ceremony, decided to indefinitely postpone the signing of the Frontiers of AI Governance Executive Order scheduled to be released on May 21. Subsequently, on May 26, Trump announced the appointment of former Attorney General Pam Bondi to the President's Council of Advisors on Science and Technology (PCAST). From the withdrawal of the executive order to the introduction of a legal heavyweight, the White House is shaping the Technology Advisory Council as the legal front line for regulatory cleanup. Bondi's litigation background, in line with Sax's deregulation advocacy, aims to use legal action to sweep away defensive technology barriers set up by various states, fully implement a laissez-faire approach to AI development at the frontier, and follow an aggressive path to defeat China.

Previously drafted by the White House, the security review draft focused on pre-market review. Although the text had diluted the initially envisioned mandatory approval regime by the Food and Drug Administration (FDA) into a "voluntary pre-market review framework," it still required developers to share models with the government up to 90 days in advance of publicly releasing a large model to assess national security risks. In response to the proposed review period, frontier labs had strongly lobbied to compress the time to 14 days before release; however, at the eleventh hour, Trump chose to withdraw the entire draft, stating dissatisfaction with the draft terms and concerns that the Federal Pre-Market Review Board would weaken the U.S.'s competitive edge against China, causing months of efforts by agencies such as the Office of the National Cyber Director (ONCD) to be completely in vain.

The White House's abrupt shift in deregulation policy aligns with the policy stance publicly disclosed by Co-Chairman of the Advisory Council, David Sacks. In response to Pope Leo XIV's encyclical Magnificent Humanity released on May 25 warning about the "threat of the AI revolution to human dignity," while Sacks agrees that AI should serve human dignity, he vehemently opposes giving the government a safety net of control, insisting that handing over control of model development to the government is the "real alignment issue." Sacks emphasizes that government control is highly likely to evolve into a tool of authoritarianism involving censorship, surveillance, and control of citizens and counters this by invoking the ancient Roman adages "Who will guard the guards themselves" and "Power corrupts, absolute power corrupts absolutely." Sacks expresses full support for former Attorney General Bondi's joining the advisory council, signaling the White House's intention to use legal actions to counteract conflicting state laws, pursue an extreme deregulation policy, and ensure that the U.S. maintains an absolute leadership position against China in the great power AI competition.