US Orders Pre-Release Review of Frontier AI in June; OpenAI Proposes Federal Framework

Frontier AI, the class of high-capability general-purpose models, has emerged in 2026 as both a driver of industrial transformation and a source of national-security risk, becoming a top issue in international politics and governance centered on the US-China rivalry. In June 2026 the US government issued an executive order requiring pre-release review of the most advanced models; the same month OpenAI published a blueprint outlining a federal framework, and Japan's financial authorities in May urged financial institutions to prepare, as responses multiply across countries and sectors.