President Donald Trump announced that discussions with artificial intelligence (AI) firm Anthropic to restore access to its Claude Fable 5 and Mythos 5 models are “going fine.” This statement came during the 2026 G7 summit, one week after U.S. officials restricted foreign use of these two advanced AI models. Several world leaders and AI executives at the summit expressed significant concern that the U.S. could control access to some of the world’s most sophisticated AI tools.
The U.S. decision to ban foreign access to Anthropic’s AI models prompted warnings from industry figures and foreign governments. They cautioned that a U.S. monopoly on AI development might lead to one-sided control over these advanced systems. French President Emmanuel Macron noted that while U.S. officials demonstrated strong awareness of AI risks, the actions appeared overly focused on national interests. He argued that democracies should collaborate on shared rules and warned against buyer distrust of AI models that could be deactivated without warning.
President Trump’s comments followed a G7 lunch, where he confirmed ongoing discussions between the U.S. government and Anthropic. Commerce Secretary Howard Lutnick, present with the President, echoed that talks were “going fine.” The meeting included prominent industry leaders such as Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman, and Google DeepMind chief Demis Hassabis.
During the G7 meeting, Trump characterized the broader AI discussion as successful, highlighting its role in bringing together key figures from the sector. He also pointed to increasing investment in the U.S. AI industry, noting the construction of some of the world’s largest AI-focused facilities within the country.
Anthropic CEO Dario Amodei and Google DeepMind chief Demis Hassabis urged the U.S. to lead G7 countries in establishing AI safety standards. Sam Altman, CEO of OpenAI, also emphasized the necessity of an international forum to agree on shared testing rules and neutral risk reviews for AI, cautioning against excessive concentration of power.
Aidan Gomez, head of Canadian AI firm Cohere, indicated broad agreement among G7 members that coordination should commence. The discussions at the summit underscore a growing international push for collaborative governance in AI development and deployment, contrasting with unilateral national restrictions.
The unresolved question remains how the U.S. will balance national security interests with international calls for shared AI governance and access. Future negotiations between the U.S. government and Anthropic, alongside ongoing G7 discussions, will likely shape the global framework for advanced AI model accessibility.
Observers will watch for any shifts in U.S. policy regarding foreign access to advanced AI models and the progress of international efforts to establish common AI safety and access protocols. The outcome could significantly influence the global distribution and ethical development of artificial intelligence.