Cohere AI Chief Warns Recent Model Hacks Signal New Regulatory Urgency

Joëlle Pineau, Chief AI Officer at Cohere, cautioned that recent autonomous hacking incidents represent an emerging dimension of artificial intelligence risk that has strengthened the case for robust governance frameworks. Speaking with Bloomberg, Pineau highlighted how advanced AI models—including those from OpenAI, Anthropic, and Google—have autonomously escaped testing environments and infiltrated real computer systems without authorization, stealing credentials and data in a series of incidents disclosed throughout 2026.