U.S. President Donald Trump said he no longer sees Anthropic, the developer of the "Mythos" artificial intelligence (AI) model, as a national security threat.
According to U.S. online outlet Axios on the 19th (local time), Trump, asked whether he views Anthropic and Dario Amodei, Anthropic's chief executive officer (CEO), as a national security threat, said, "A week ago, maybe, but not now."
He added, "I was in the G7 meeting room with (CEO Amodei), and he seemed like a good person. A smart guy," and said, "(After the export control guidance) he responded to us very quickly. I think he handled it very responsibly, and it looks like he'll do well going forward."
On the 12th, the U.S. Commerce Department ordered an export ban blocking foreign access to Anthropic's top-tier, high-performance AI models, Claude "Mythos 5" and "Fable 5," citing national security. The move was reportedly driven by concerns that military and intelligence figures in adversary nations such as China and Russia could use the models.
However, Trump appeared to soften his stance after meeting Amodei at the G7. At the meeting, Amodei was quoted as telling Trump, "We want the U.S. government to lead global AI governance."
Describing the situation at the time, Trump said, "It was an issue carrying enormous legal liability, but Amodei responded very quickly." He continued, "People sometimes go straight to prison over things like that. You can't take that lightly," adding, "I think CEO Amodei handled it very responsibly."