White House AI Guidelines Exempt U.S. Open Models From Government ReviewThe Wall Street Journal ^ | Updated Aug. 4, 2026 7:07 pm ET | Amrith RamkumarPosted on by E. Pluribus UnumThe new framework, completed in recent days, is seen as most likely to affect OpenAI, Anthropic and GoogleWASHINGTON—The Trump administration’s new guidelines for powerful artificial-intelligence tools exempt open models made by U.S. companies, a gap that marks the White House’s latest bid to ensure AI tools are safe without slowing down innovation. TOPICS:Computers/InternetKEYWORDS:To: E. Pluribus UnumSummary of Article The Trump administration’s new voluntary AI guidelines exempt open models developed by U.S. companies from government review, while applying scrutiny mainly to certain closed, proprietary models. The framework, completed in recent days and discussed with executives from leading AI companies on Tuesday (August 4, 2026), aims to promote AI safety without hindering innovation. Open models (also referred to as open-weight models) from U.S. firms are carved out of the review process. The guidelines stem from a June 2026 executive order establishing a voluntary process for assessing advanced cyber risks of frontier AI models, with government access for up to 30 days before release under confidentiality protections.