Anthropic to offer EU access to its advanced Mythos model

Source Domain: www.cnbc.com

Anthropic has agreed to share access to its advanced AI model Mythos with the European Union (EU) following months of discussions due to cybersecurity concerns.
The European Commission expressed its approval of the developments, emphasizing the need for a clearer understanding of the potential risks of such powerful AI technology.
Mythos, launched in April as part of Anthropic’s Project Glasswing, is excellent at identifying security flaws in software, prompting concerns about increased cybercrime risks if misused.
The EU aims to address the potential shared challenge of advanced AI models and has intensified discussions with like-minded partners, including the United States.
Official terms and conditions of the agreement have not been confirmed; Anthropic did not immediately respond to CNBC’s request for comment.
Governments, banks, and tech firms initially raised alarms over Mythos’s capabilities when it was first released, fearing its potential misuse by bad actors.
The U.S. government had previously restricted AI model access to non-U.S. governments as part of its strategy to maintain dominance in the AI field.

You may have missed