Jamie Dimon: Anthropic AI Risks a "Real Issue

JPMorgan Chase CEO Jamie Dimon warns that risks from Anthropic's Mythos AI model are a "real issue." He believes U.S. government officials are now focused on these potential risks. Dimon raised concerns about broadly distributing powerful AI systems. He compared widespread access to Mythos to "giving ballistic missiles to individuals." This statement was made during a defense and innovation summit. Anthropic is at the center of a debate about managing cybersecurity risks from powerful AI models. They launched Project Glasswing, a defensive cybersecurity initiative using an advanced, unreleased "Claude Mythos" AI model. Instead of a public release, Anthropic initially partnered with over 50 organizations, including Google, Microsoft, and Apple. The goal is to proactively scan and patch software vulnerabilities before adversarial AI can exploit them. Anthropic has since expanded access to approximately 150 new organizations. The company stated that Project Glasswing and the capabilities of Claude Mythos Preview have sparked broad conversations about how AI is changing cybersecurity. Their role, they say, is to help the software industry adapt by safely providing wide access to better models and tools. They also aim to shift support from finding vulnerabilities to disclosing, fixing, and deploying patched software. This highlights ongoing concerns about the safe development and deployment of advanced AI technologies.