Anthropic suspends a new model over misuse concerns as the White House objects
Anthropic suspended access to a newly released model after concerns that it could be misused for cybersecurity attacks and hacking, according to the BBC. The decision pulled a capability the company had only recently shipped, signaling that its own internal review found the risks outweighed the benefits of leaving it available.
The restriction quickly became political. Defense Secretary Pete Hegseth and other White House allies intensified their criticism of the company , with the discontinuation of model access tied to a Trump administration directive. The friction underscores a widening gap between AI developers that throttle their own products on safety grounds and an administration pressing for fewer constraints.
The episode is a useful tell about where the AI-policy fight is heading: when a leading lab voluntarily withdraws a model, it invites pressure from a government that wants those capabilities to remain available.