White House executive order speeds plans for U.S. AI model standards

The Trump administration is in advanced talks with AI companies to finalize voluntary standards for frontier AI model releases after building a federal oversight framework through a series of executive orders, government review requirements and intervention in product launches. The standards, which could be announced soon, would set benchmarks for advanced models and clarify who can access them in the United States and abroad. The framework rests on three executive orders. A December 2025 order directed the Commerce Department to evaluate state AI laws, identify rules seen as conflicting with federal policy and created an AI Litigation Task Force to challenge state regulations the administration considers too burdensome. A June 2026 order tasked the National Institute of Standards and Technology with creating a classified benchmarking process for frontier models within 60 days and introduced a pre-release evaluation window of up to 30 days of government access before covered models can be shared with trusted partners. Oversight is being handled through existing agencies, mainly NIST and the Cybersecurity and Infrastructure Security Agency, rather than a new federal AI regulator. Federal intervention has already shaped product rollouts. On June 12, Commerce Secretary Howard Lutnick ordered Anthropic to block foreign access to Claude Fable 5 and Claude Mythos 5 three days after launch. Anthropic later said it could not enforce nationality-based restrictions and shut both models down for everyone before Fable 5 access was restored and limited permission was granted for Mythos 5 to more than 100 approved U.S. organizations. OpenAI also delayed a full public launch of GPT-5.6 at the government's request, while Google has been discussing advanced coding model releases with officials. The effort to preempt a patchwork of state AI rules remains incomplete because congressional action has not materialized, leaving companies to navigate both federal controls and state-level regulation as Anthropic and OpenAI prepare for IPOs. Anthropic filed a confidential S-1 with the SEC on June 1 after raising $65 billion at an implied $965 billion valuation.