JPMorgan Chase CEO Jamie Dimon is raising serious concerns about Anthropic's advanced AI model, Mythos, warning it could fundamentally reshape the threat landscape for the entire financial sector.
At a summit on July 15, Dimon compared granting broad public access to Mythos to handing out "ballistic missiles to individuals." His alarm centers on the model's proven ability to identify and exploit software vulnerabilities.
Mythos, part of Anthropic's Claude family, scored exceptionally high on cybersecurity benchmarks during a preview in April 2026. Rather than a public release, Anthropic launched "Project Glasswing," a restricted program granting access to roughly 12 core partners and over 40 other vetted organizations for defensive security purposes, backed by $100 million in usage credits.
JPMorgan is a core partner. Dimon first highlighted the risk during the bank's Q1 earnings call in April, stating Mythos creates "additional vulnerabilities" for finance. His July comments significantly escalated the rhetoric.
The debate has caught the attention of crypto firms as well. Coinbase has reportedly sought access to the model for its own security testing.
Dimon remains one of the most influential voices in traditional finance. When he frames an AI model as a systemic risk, institutional investors pay attention. Any major hack linked to AI-assisted techniques could trigger a broad risk-off reaction across digital asset markets.