ChinaAI & Machine Learning

Trump’s AI order seeks security safeguards without slowing race with China

US President Donald Trump signed a long-awaited executive order on artificial intelligence (AI) on Tuesday, which asked technology companies to voluntarily share advanced models before they are released, signalling a shift in the administration’s largely la...

Donald Trump signed a long-awaited executive order on artificial intelligence Tuesday, asking technology companies to voluntarily share advanced models before release. The directive signals a shift in an administration that had largely avoided AI regulation since returning to office. But the voluntary nature of the new framework leaves a crucial question unanswered: how much oversight is enough without ceding ground to Beijing? The order came less than two weeks after Trump postponed a more aggressive version, which reportedly proposed a mandatory 90-day review window. That earlier draft sparked fierce internal debate, with critics warning it could hobble US firms competing against Chinese rivals like Baidu and SenseTime. The final text drops the hard timeline, replacing it with a request for voluntary submissions. What changed? The rise of Anthropic’s Mythos model, Mythos, a powerful new AI system, reportedly set off alarms inside the White House over its potential for misuse in disinformation or cyberattacks. The incident forced the administration to confront a paradox: Trump’s hands-off policy had allowed US innovation to surge, but it also left Washington blind to emerging risks. The executive order directs the Treasury, Defence, Commerce, and Homeland Security departments, along with other agencies, to secure agreements with AI developers for testing. The White House framed the move as a national security necessity. “Advanced AI capabilities make our nation stronger, but also introduce new national security considerations that require coordinated action,” it said. Trump himself has been blunt about his priorities. “We’re leading China, we’re leading everybody, and I don’t want to do anything that’s going to get in the way of that lead,” he told reporters in the Oval Office in May. The new order reflects that calculus: it buys Washington a window into frontier models without imposing binding constraints that could slow development. The voluntary nature of the framework means US firms still set their own pace. Companies like OpenAI, Google DeepMind, and Anthropic can choose what to share and when. This leaves a gap: Chinese firms, operating under Beijing’s centralized AI strategy, face no such transparency requirements. Washington’s oversight is essentially a one-way mirror—it sees US labs, but not their competitors. A casual reader might miss the deeper tension here. Trump’s order reverses his own first-day action, which scrapped a Biden-era AI oversight order that also relied on voluntary commitments. The administration has come full circle, but with a crucial difference: the earlier framework required safety test results to be shared. The new one simply asks for models to be submitted for testing, with no mandated disclosure of results. The real test will come when the next Mythos-class model emerges. If a US company declines to share it, or if a Chinese firm leapfrogs American capabilities while Washington watches from the sidelines, the voluntary framework could face intense pressure to harden into regulation. For now, the race with China continues—and the US government is betting that asking nicely is enough to stay ahead.

US President Donald Trump signed a long-awaited executive order on artificial intelligence (AI) on Tuesday, which asked technology companies to voluntarily share advanced models before they are released, signalling a shift in the administration’s largely la...

Trump’s voluntary AI vetting buys Washington oversight without imposing hard limits, but leaves US firms racing China still setting their own pace.

The development adds to a wider China ai & machine learning story in which companies are being judged on execution, capital access, regulatory fit and the credibility of their regional expansion plans.

For business readers, the important question is whether this becomes an isolated announcement or part of a more durable operating pattern across customers, financing channels, partners and public-market expectations.

Related News

More from this beat

Greater China / AI & Machine LearningAI PCs to make up more than half of sales in 2026, says Asus coChina / AI & Machine LearningMore US firms turn to China’s DeepSeek over pricey Silicon Valley AIChina / AI & Machine LearningA Chinese robotics start-up beat Nvidia on a global AI ranking. Is a new tech war brewingGreater China / AI & Machine LearningYip In Tsoi banking on major rejig