EU gains access to anthropic's mythos 5 and openai's gpt-6-astra models

The European Union's cybersecurity agency, ENISA, has been granted access to Anthropic's Mythos 5 and OpenAI's GPT-6-Astra models, according to a statement from the European Commission. The access is currently in a testing phase. The Commission confirmed the development on September 10, 2026, following reports from Reuters. This follows a policy shift by the U.S. White House, which initially restricted foreign entities from accessing high-performance AI models but later eased those restrictions. Despite the change, the eligibility of foreign institutions to access such models remained unclear. The move marks a significant step in international collaboration on AI security, with ENISA now able to evaluate these advanced systems for potential threats.