The White House is in advanced discussions with leading artificial intelligence companies over voluntary standards and release timelines for new models, according to sources cited by the Financial Times. The guidance could be announced as soon as next week, following government intervention in recent rollouts by Anthropic and OpenAI.
The talks reflect growing urgency in Washington to establish guardrails for the rapidly evolving AI industry without resorting to formal legislation. By focusing on voluntary commitments, the administration aims to preempt safety concerns while avoiding the political gridlock that has stalled broader tech regulation.
The initiative comes after the White House previously engaged with firms like OpenAI and Anthropic on model testing and transparency. Specific terms under discussion include pre-release safety assessments and phased deployment schedules, though details remain sparse as negotiations continue.
If finalized, the standards would mark the most concrete federal framework for AI governance to date, potentially influencing global norms. Critics argue voluntary measures lack enforcement teeth, while industry insiders worry about compliance costs and competitive disadvantages.
Some experts caution that without statutory authority, such guidelines could prove toothless against lapses in safety or fairness.