Jamie Dimon says broad access to Mythos would be like giving 'ballistic missiles to individuals' (www.businessinsider.com)

JPMorgan Chase CEO Jamie Dimon raised concerns about the potential risks associated with broad access to Anthropic's Mythos-class AI models during the Pennsylvania Defense and Innovation Summit. He likened access to these powerful models to giving "ballistic missiles to individuals," highlighting the significant cybersecurity threats they pose. Dimon's remarks underscore the importance of government oversight as the US seeks to ensure that such advanced technologies, particularly those capable of identifying critical vulnerabilities in operating systems, are managed responsibly. Anthropic's Mythos 5 is its most advanced model, previously deemed too effective at detecting "high-severity vulnerabilities," leading the company to restrict access to a select group of US organizations. After an initial attempt to release Fable 5 to the public, which was blocked due to concerns about its security guardrails, the US government imposed export controls, preventing foreign users from accessing these models. Ultimately, Anthropic announced that these restrictions had been lifted, indicating a cautious return to wider availability while emphasizing the need for safeguards in deploying such powerful AI technologies.