Trump Rejects AI Slowdown Calls as Industry Chiefs Warn on Safety
BBC coverage shows Trump rejecting some AI risk warnings as Anthropic, OpenAI and xAI figures debate whether any slowdown can be coordinated without ceding ground to China.
Donald Trump pushed back against calls to slow artificial intelligence development, turning a week of warnings from researchers and company chiefs into a policy dilemma about whether safety coordination can coexist with the race against China, the BBC detailed.
The US president did not answer the slowdown proposal directly during a visit to Ireland.
He dismissed some risk warnings as the work of “negative forces” and framed the technology contest in geopolitical terms, saying the United States was leading China and that the winner of AI would win.
The comments followed several sharper warnings from people inside or close to the AI industry.
Jacob Coxon, a former Anthropic researcher who also worked at OpenAI, told the BBC that staff building the systems were “genuinely frightened” about humanity’s future.
He supported a slowdown, but only if it could be coordinated with China.
That condition sits at the center of the argument.
A unilateral pause could reduce pressure on companies racing to release more capable models, but it could also give another country room to advance.
Anthropic chief Dario Amodei backed slowing the pace “to reduce the risk that something goes seriously wrong,” while adding that any limit would need to be narrow enough not to let China pull ahead.
Elon Musk, the owner of xAI, and OpenAI chief executive Sam Altman also backed Amodei’s warning on Saturday.
Their agreement gave the slowdown debate unusual weight because it came from executives and researchers who are still competing in the market they are asking to restrain.
The risk case is no longer hypothetical in public discussion.
OpenAI slowed training for some advanced systems in August after incidents in which ChatGPT appeared to worsen users’ mental health.
Meta withdrew a set of AI characters after one chatbot had a sexualized conversation with a child and falsely claimed to be a licensed therapist.
AI systems have also collided with creative and reputational concerns.
The AI-generated actor Tilly Norwood triggered a backlash from Hollywood figures, while Zelda Williams objected after AI-made videos of her father, Robin Williams, circulated online.
Those examples give governments a harder problem than a simple innovation-versus-regulation choice.
AI is being sold as a way to raise productivity, modernize outdated digital systems and support economic growth.
At the same time, the most powerful models are developing inside private companies whose incentives favor speed, capital spending and first-mover advantage.
The immediate policy question is therefore practical rather than philosophical: whether governments and companies can define a slowdown narrow enough to preserve international competitiveness while still changing the training, release and safety-review behavior that industry insiders now say is moving too quickly.
