BNP Paribas, Europe's largest bank by assets, is teaming up with French AI company Mistral AI to counter a new wave of cybersecurity threats driven by advanced AI systems. The move comes after Anthropic's Mythos model demonstrated a 72% exploit success rate and identified a bug in OpenBSD that had gone undetected for 27 years.
The collaboration, which began in September 2023 with a pilot in BNP's Global Markets division, expanded into a multi-year agreement in July 2024. The latest phase, announced May 26, focuses on developing a dedicated cybersecurity model for European banks.
BNP Paribas CIO Marc Camus stated the goal is to anticipate threats from any advanced AI that could be weaponized against financial infrastructure. Mistral AI is also in talks with other European banks to deploy this technology.
This partnership reflects European institutions' preference for regional AI developers, reducing data compliance risks by keeping infrastructure within European jurisdiction. For investors, the shift is clear: traditional hacking defenses are insufficient against AI-powered attacks operating at machine speed and scale. Mistral AI is positioning itself as a key European alternative, potentially creating a network effect that strengthens defenses across the continent.