The administration of US President Donald Trump has reversed an earlier decision that restricted American artificial intelligence companies from sharing their new models with the British government before their official release, allowing British authorities to test the models and assess their safety.
اضافة اعلان
The British newspaper The Telegraph reported on Sunday that US AI developers can now grant the British government early access to their models for safety testing without violating the Trump administration’s policy.
According to Reuters, the US administration has reversed previous guidance that instructed companies, including OpenAI and Anthropic, not to share their new models with the UK AI Safety Institute before launch.
The shift signals a more flexible approach by the Trump administration toward international cooperation on AI safety, giving American companies greater scope to work with Britain as governments face growing challenges in regulating and overseeing rapidly evolving technologies.
Established following the first global AI Safety Summit in 2023, the UK AI Safety Institute is among the bodies responsible for testing advanced models, assessing their technical capabilities and identifying potential risks associated with their misuse.
The US reversal comes despite Trump’s previous opposition to imposing mandatory safety requirements on AI developers. He has repeatedly downplayed calls for stricter binding regulations on companies operating in the sector.
However, recent developments suggest a possible gradual shift in this approach, particularly after a newly launched US task force focused on AI announced on Friday that developers should be required to report security incidents involving their models.
Meanwhile, spokespersons for the US president, OpenAI and Anthropic did not immediately respond to requests for comment from Reuters.
The decision comes amid growing international debate over how to oversee advanced AI models and how to balance accelerating technological innovation with stronger safety measures and efforts to limit the risks of misuse.
Resource: Al-Ghad.
