Trump's AI Safety Push Leaves Crypto-Native Projects in Regulatory Limbo

At the G7 Summit in Évian-les-Bains, France, President Trump hosted a working lunch to discuss AI safety testing frameworks. The event brought together CEOs from Google DeepMind, OpenAI, and Anthropic with G7 heads of state. Demis Hassabis of Google DeepMind was a key participant.

The conversation centered on establishing global norms for deploying advanced AI models safely and quickly. Trump's 'Winning the Race: America's AI Action Plan,' published in July 2025, laid out a framework prioritizing innovation, infrastructure investment, and international diplomacy over prescriptive regulation.

Voluntary safety testing frameworks have been initiated by the administration, with Google, OpenAI, Anthropic, and Meta participating. No mandates or penalties are involved, just voluntary participation. This means established AI companies like these can deploy their models faster with fewer bureaucratic obstacles.