Hacker group ShinyHunters claims to have infiltrated Anthropic's internal systems, according to a message relayed by Dark Web Intelligence. The group reportedly accessed sensitive areas related to Anthropic's Mythos model, including user management panels and AI experiment dashboards. Screenshots of model performance metrics and cost analyses were also shared. Anthropic has not yet confirmed the authenticity of these claims. If verified, the breach could pose security risks to major tech firms and cryptocurrency companies that have trialed the model.