US technology company Anthropic has offered the European Union Agency for Cybersecurity (ENISA) access to its latest artificial intelligence model, Mythos. Discussions now concern the arrangements needed to guarantee “secure access” on the European side, a source close to the matter confirmed to Agence Europe.
“Our priority now is to continue working with Anthropic in order to agree on terms and conditions for future access. We must put in place an appropriate framework guaranteeing secure access to the model”, this source explained.
Officially, the European Commission and ENISA have confined themselves to welcoming “the latest developments on potential future access” following “several productive meetings with Anthropic”. “But this has not yet been confirmed”, an ENISA source told Agence Europe. “We welcome the latest developments on potential future access”, European Commission spokesperson Thomas Regnier said, while stressing the importance of having “a clear picture on the potential risks” associated with this model. “Let’s not forget that Mythos is not one off. A new wave of powerful models are coming to the market. This is a shared challenge, and we are intensifying our discussions with like-minded partners, including the United States”, he added.
From 2 August, under the European regulation on artificial intelligence (AI Act), the European Commission will have full enforcement powers. It will, in particular, be able to demand access to AI models and legally compel providers to implement risk mitigation measures for their most advanced artificial intelligence models. (Original version in French by Ana Pisonero Hernández)