OpenAI, Google, Anthropic: Safety Talks Nobody Told Trump
The White House position is that safety discourse slows American AI development at a moment China cannot afford to be given ground.
OpenAI, Google, Anthropic: Safety Talks Nobody Told Trump
OpenAI, Anthropic, and Google DeepMind have been conducting weeks of closed-door AI safety talks — a coordinated effort that places the three most powerful AI laboratories in direct tension with the White House, according to TechCrunch. The conversations, confirmed by OpenAI, focus on shared safety standards at a moment when the Trump administration has publicly dismissed AI risk concerns as a distraction from competition with China.
The timing is pointed. At a separate summit, President Trump interrupted a live event to call Nvidia chief Jensen Huang directly, declaring that concerns about artificial intelligence are a "hoax" — a claim that landed as the three companies were, by their own account, meeting specifically to address those concerns.
The White House position is that safety discourse slows American AI development at a moment China cannot afford to be given ground. The labs appear to have reached a different conclusion: that coordination is the minimum viable defence against systems neither government nor industry fully understands yet.
What happens next carries consequence well beyond Silicon Valley. The gap between what the administration believes about AI risk and what its own leading developers believe is now a matter of public record — and it is not a small gap. For anyone tracking where regulation in this space is actually heading, the lobbying battle that follows these talks will be the one to watch.
*— Isla Camilleri, Global Affairs & Lifestyle Editor*
