United States CEO warns Anthropic AI poses security risks | JQJO

PENNSYLVANIA – JPMorgan Chase CEO Jamie Dimon warned that Anthropic’s advanced Mythos artificial intelligence model poses serious security risks if released broadly to the public. Speaking on Wednesday at the Senator Dave McCormick Defense and Innovation Summit in Pennsylvania, Dimon said that allowing individuals to access Mythos would be comparable to putting ballistic missiles in their hands, underscoring the model’s ability to identify software vulnerabilities. He described the system’s capabilities as a “real problem” and noted that the U.S. government is already paying close attention to the potential dangers associated with such powerful technology.

UNITED STATES – Dimon’s comments come after Anthropic itself disclosed that Mythos is so capable at discovering software flaws that the company believes it cannot safely make the model publicly available. That decision has intensified formal reviews and risk assessments by both industry stakeholders and policymakers, who are examining how advanced AI models might be misused if widely deployed. The debate around Mythos reflects broader concerns about the rapid development of cutting-edge AI systems and the need to evaluate their security implications, as government agencies and private-sector leaders focus on preventing malicious exploitation of these tools.