At the recent G7 Summit, international leaders voiced growing alarm over the United States capacity to unilaterally restrict access to advanced artificial intelligence models. French President Emmanuel Macron and Indian Prime Minister Narendra Modi warned during high level discussions that sudden U.S. export controls could destabilize global economies and compromise critical infrastructure. The concerns were catalyzed by the Trump administration recent decision to block the export of Anthropic newest Mythos 5 and Fable 5 models, citing national security risks after Amazon alerted the White House to potentially bypassable safety guardrails. Although cybersecurity analysts noted that comparable capabilities exist in openly available models, the suspension underscores a systemic vulnerability: governments and enterprises building on American AI infrastructure face the risk of overnight access revocation without transparency. Leaders emphasized that democratic nations require guaranteed, uninterrupted access to foundational AI systems to safeguard economic resilience and national security. Aidan Gomez, CEO of Canadian AI firm Cohere, echoed this sentiment, arguing that reliance on a narrow cohort of American tech giants poses a direct threat to digital sovereignty and long term economic independence. The incident has accelerated diplomatic efforts to decouple global AI development from unilateral U.S. export policy. During summit proceedings, delegates explored a trusted partners framework designed to grant non American nations and vetted companies continued access to models from providers like Anthropic and OpenAI. The proposal aims to establish a coordinated trade network that aligns model usage with allied security objectives, including strategic competition with China, while mitigating sudden supply disruptions. Despite the proposal potential, its practical implementation remains uncertain. Industry observers question whether such a framework would adequately protect startups and mid sized enterprises that lack the diplomatic leverage to secure guaranteed access. The debate highlights a broader geopolitical tension: while European and non Western nations continue to advocate for AI sovereignty to reduce foreign dependency, American models maintain a decisive performance lead, making complete technological separation economically impractical. As the United States tightens export oversight, allied governments are increasingly weighing the trade offs between security compliance and the operational certainty required to sustain global AI innovation. The G7 ongoing negotiations will likely set the precedent for how advanced AI infrastructure is governed, distributed, and protected from geopolitical volatility in the coming years.