Jamie Dimon says broad access to Mythos would be like giving 'ballistic missiles to individuals'

At the Pennsylvania Defense and Innovation Summit, JPMorgan Chase CEO Jamie Dimon raised significant concerns regarding the potential dangers of unrestricted access to Anthropic's cutting-edge Mythos-class AI models. He emphasized that such access could be likened to providing individuals with 'ballistic missiles.' During the summit, which featured notable attendees including Pennsylvania Senator Dave McCormick and other political and defense leaders, Dimon stated that the U.S. government is diligently monitoring the situation. He highlighted the importance of ensuring that the AI models remain controllable upon their release, labeling the situation as a 'real issue.' Mythos 5, recognized as Anthropic's most advanced AI model, has the capability to identify vulnerabilities within operating systems. However, both critics and Anthropic itself have expressed serious concerns about the cybersecurity threats it could pose if made widely available. In April, Anthropic decided to halt the model's release, citing its proficiency in uncovering high-severity vulnerabilities. Rather than offering broad access, the company opted to restrict availability to a select group of U.S. organizations. In June, Anthropic introduced Fable 5 to the public, a model designed with safety measures to mitigate risks related to cybersecurity and other sensitive topics. However, shortly after its release, the U.S. government intervened, blocking access to both Mythos 5 and Fable 5 due to concerns that the established guardrails might be circumvented. This led to the imposition of export controls to prevent foreign access. In response, Anthropic temporarily suspended access for all users. On June 30, the AI firm announced that it had received notification from the Department of Commerce lifting the export controls, allowing them to begin restoring access to their models. They expressed gratitude to users for their patience and promised to provide updates on the redeployment process. As the debate around AI safety continues, both Anthropic and JPMorgan Chase have yet to comment on the developments.