San Francisco: OpenAI chief Sam Altman and Tesla and xAI chief Elon Musk have teased powerful new AI models just days after publicly agreeing with Anthropic CEO Dario Amodei that the pace of artificial intelligence development should slow down.

The apparent about-turn has reignited one of the tech industry’s fiercest debates: whether AI labs are moving too fast for safety, or whether slowing down would simply hand the advantage to rivals.

Amodei has been among the most prominent voices calling for caution, arguing that increasingly capable AI systems need stronger safety testing before release. Altman and Musk had seemed to nod along, only to return days later with hints of major upcoming launches.

Talk of caution, race of ambition

The contradiction captures the central tension of the AI industry. Every leading lab publicly acknowledges the risks of powerful AI, from misuse to unforeseen behaviour, while privately racing to ship the next breakthrough before competitors do.

Investors, talent, and market share all flow to whoever leads the frontier. That commercial pressure makes voluntary slowdowns difficult to sustain, as each lab fears that pausing would let others pull ahead.

What comes next

Regulators in the United States, the European Union, and elsewhere are watching closely. The EU’s AI Act and various proposed American rules aim to impose testing and transparency requirements on the most powerful models, but enforcement remains a work in progress.

For now, the pattern looks set to continue: public calls for responsibility paired with private acceleration. Until either regulation or a genuine industry consensus changes the incentives, the AI race will keep running at full speed, whatever its leaders say about slowing down. Industry watchers note that the gap between public pledges and private behaviour is now the norm in the AI race. Until safety rules carry real enforcement power, few expect the leading labs to voluntarily ease off the accelerator.