French AI startup Mistral AI is reportedly discussing with European banks how to deploy an alternative to Anthropic PBC’s “Mythos,” a limited-access AI model aimed at identifying cybersecurity vulnerabilities at high speed and scale. According to Bloomberg Technology and PYMNTS, Mistral is working on an answer to Mythos and is engaging banks that do not have access to Mythos directly. The reports say Mistral’s discussions are focused on enabling banks to use an AI-based tool for cybersecurity testing or vulnerability discovery without relying on the American model. Both sources describe the effort as being in talks and based on information from people familiar with the matter. The reports do not provide details on timelines, pricing, specific technical capabilities beyond the general goal of uncovering vulnerabilities at speed and scale, or which banks are involved. The disclosures also do not indicate whether Mistral’s model will be delivered as a hosted service, offered through partnerships, or integrated into existing bank security workflows.