Trump stated that he no longer considers Anthropic a national security risk following his G7 meeting with the CEO.
TL;DR: Trump told Axios that Anthropic has “behaved very responsibly” and indicated he might relax restrictions on its Fable 5 and Mythos 5 AI models. In a pre-recorded interview with Axios, President Donald Trump stated he no longer considers Anthropic a national security threat, a significant change from the previous aggressive stance of his administration towards the AI company over the last three months. When asked if he sees Anthropic as a threat, Trump responded, “Well, not now. But a week ago, maybe,” adding that the company has “behaved very responsibly.”
These remarks come shortly after the Commerce Department issued a directive on June 12, requiring Anthropic to obtain US government approval before allowing foreign nationals to access its Fable 5 and Mythos 5 models, the company's most advanced AI systems. This order followed months of increasing tension between the administration and Anthropic due to the company's refusal to remove specific safety measures from its military products, which led to crisis-level discussions between Anthropic and Commerce officials last week.
Trump met with Anthropic CEO Dario Amodei on Wednesday during the G7 Summit in Évian-les-Bains, France, an interaction that seems to have influenced the president’s viewpoint. This meeting followed earlier discussions between Anthropic's senior technical staff and Trump administration officials earlier in the week. Trump told Axios that he would consider easing the restrictions, stating, “I would, but I’m not sure I have to do that,” in response to a question about potential rollbacks.
The origins of the dispute date back to March 2026 when the Pentagon identified Anthropic as a supply-chain risk after the company declined to remove guardrails associated with surveillance and autonomous weaponry from military products. Commerce Secretary Howard Lutnick subsequently sent a letter threatening criminal charges against the company, which drew criticism from tech industry groups and prompted lobbying from allied governments, including the UK, for exemptions.
The timing of Trump's conciliatory approach is crucial as Anthropic confidentially filed for an initial public offering in early June, with a valuation reported by Fortune at around $965 billion. The existing federal restrictions had created uncertainty regarding the IPO, and any indication of de-escalation from the White House could help stabilize investor confidence prior to the offering.
Trump characterized the situation as creating “tremendous liability” for the administration, acknowledging that the crackdown had faced backlash from both industry and allies. While he stated he would not shut down Anthropic, he did not commit to a specific timeline for lifting the Commerce Department's directive.
This shift does not eliminate the underlying conflict, as the Pentagon’s supply-chain designation still stands, and the June 12 order from the Commerce Department has not been officially rescinded. Anthropic has not publicly indicated whether it intends to change its guardrail policies to meet military demands.
What has changed is the political signal from the top: Trump seems open to negotiation rather than escalation. Amodei has been exploring various avenues to resolve the standoff. At the G7 summit, he and Google DeepMind CEO Demis Hassabis jointly proposed a US-led AI coalition to G7 leaders, framing Anthropic as a collaborative partner in American technology diplomacy rather than a regulatory adversary. This approach seems to have provided Amodei with direct access to Trump during a time when the president was more amenable.
Whether these positive comments will lead to policy changes remains uncertain. The Commerce Department operates with significant autonomy concerning export controls, and reversing a formal directive involves bureaucratic processes that a single interview cannot expedite. For Anthropic, the Axios interview represents a political victory, but the legal and regulatory constraints will persist until the administration takes action.
Other articles
Trump stated that he no longer considers Anthropic a national security risk following his G7 meeting with the CEO.
Trump informed Axios that Anthropic has "acted very responsibly" and indicated that he might relax restrictions on its Fable 5 and Mythos 5 AI models.
