WASHINGTON, October 12, 2026 — The United States has reportedly softened its position on international artificial intelligence safety testing, potentially allowing American AI developers to give British researchers early access to advanced models before their public release.

According to a Reuters report citing the British newspaper The Telegraph, the Trump administration has reversed an earlier position that discouraged companies such as OpenAI and Anthropic from sharing new AI models with Britain's AI Security Institute.

The reported change comes as governments and technology companies face growing pressure to assess the cybersecurity, misuse and safety risks associated with increasingly capable AI systems.

Why AI Safety Testing Matters

The UK's AI Security Institute evaluates advanced AI models to understand their capabilities and potential risks. Early access can help researchers test systems before wider public deployment, potentially identifying vulnerabilities that might otherwise go unnoticed.

For American technology companies, international cooperation could support more consistent testing practices and improve communication between AI developers and government researchers.

However, the policy shift does not mean that all AI models will automatically be available to foreign researchers. Access arrangements may depend on individual companies, security requirements and government guidance.

What It Means for the US Technology Industry

The reported change could influence the future of AI safety collaboration between the United States and the United Kingdom. It also highlights the challenge facing policymakers as they balance technological innovation, national security and responsible AI development.

Representatives for the Trump administration and the companies named in the report did not immediately respond to Reuters' requests for comment.

Further clarification from officials will be important in determining how the reported policy change will be implemented and what it means for the next generation of AI systems.