OpenAI will hand over its newest AI models to U.S. regulators a month early, after President Trump signed an executive order demanding early access. The order, softened from a 90‑day request to 30 days, follows backlash and Trump’s own doubts. OpenAI’s compliance signals a shift in how tech firms manage government oversight.

George Osborne, OpenAI’s head of countries, said the company will submit its next‑gen models for a benchmarking process that assesses “advanced cyber capabilities.” The test will help define which models qualify as a ‘covered frontier model,’ a category that could trigger stricter regulation. This framework aims to keep emerging AI tools in check for development.

Osborne highlighted OpenAI’s proactive proposals to help governments monitor AI safety beyond the U.S. He spoke at SXSW in London, urging broader international cooperation. By sharing data ahead of release, OpenAI hopes to set a precedent for responsible deployment. The move could pressure other firms to adopt similar transparency measures in the global tech landscape.

The voluntary compliance may ease tensions between AI innovators and regulators. It also clarifies the scope of the executive order, limiting mandatory disclosures to a one‑month window. For consumers, earlier access to vetted models could mean safer, more reliable tools. The policy could redefine industry standards for pre‑release AI evaluation in the technological future sector.