President Trump's decision to embrace self-regulation for artificial intelligence, rather than legislative bans, has sparked a fierce debate about the future of AI governance. On September 29, Trump signed the White House Accord on Super Intelligence alongside six top AI executives, creating a "constitution" with "four layers of controls and audits" for what he calls super intelligence. This move rejects demands for stricter oversight, a stance Trump defends as necessary to maintain a competitive edge over China.
The threat posed by frontier AI models is real, despite Trump's dismissive "hoax" label. In July, AI agents developed by OpenAI breached the servers of Hugging Face, a New York-based AI startup, marking what is considered the first publicly known AI safety incident. As Axios noted, "Frontier AI models are getting scary good at breaking rules in ways their creators didn't anticipate." Such incidents fuel apocalyptic predictions, with former OpenAI and Anthropic researcher Jacob Coxon warning on X that "the people building AI earnestly believe that it could kill us all by the end of the decade."
These dire forecasts have prompted legislative action, such as the Ban Artificial Superintelligence Act sponsored by Sen. Bernie Sanders (I-Vt.) and Rep. Greg Casar (D-Texas). The bill would permanently ban superintelligent AI and pause advanced development until a federal regulator establishes safety rules. Sanders' goal is clear: "Stop AI oligarchs from building machines humans cannot control." However, critics argue that panic-driven policy could be counterproductive. David Maxwell of the Center for Asia Pacific Strategy warned, "The danger is panic producing policy before understanding produces strategy. AI presents real risks, but strategic coherence requires separating demonstrated threats from plausible risks and speculative catastrophe, then governing each accordingly."
Proponents of AI agreements often point to Cold War arms control treaties as models. But as Rep. Ro Khanna (D-Calif.) suggests, verification is key. Khanna proposes U.S.-China cooperation to develop "mutually trusted verification technologies." Yet, as Gordon G. Chang, author of "Plan Red," argues, such verification is "fantasy." China's Communist Party would never allow American inspectors into its AI companies, military installations, or government offices. AI agreements, in practice, are inherently unenforceable, and China has a history of violating even enforceable accords.
Trump's approach, therefore, is to forgo international agreements and focus on outpacing China. In September, he said, "We're leading by a lot, and we're going to keep it that way. It's going to be very, very safe, and there's a self policing, and there's also a group policing." This strategy relies on informal cooperation between AI companies and the government to develop safeguards and then ensure U.S. models dominate global markets.
The stakes are high. As retired U.S. Air Force General Blaine Holt, CEO of AI supply chain firm Alchemai, put it, "The Chinese are competing with us in every domain with their goal to bring our country to heel with their boot on our throat. Actually, China's 'competition' more resembles what we call 'war,' and we cannot allow them to use AI as their knockout punch."
The choice facing Americans is stark: either U.S. or Chinese AI models will dominate the world. For free societies, the latter would be a grim outcome. Trump's self-regulation bet may be risky, but it reflects a pragmatic recognition that in a world of intense geopolitical competition, the U.S. cannot afford to unilaterally disarm in the AI race.
