Canada's Banking Regulator Warns Banks About Anthropic's Claude Mythos AI Model
July 14, 2026 - 7:11 am
Image by: Canva
An April email to bank technology chiefs, released under access-to-information rules, highlights the growing concern among regulators about the rapid advancement of artificial intelligence. Specifically, it mentions Anthropic’s Claude Mythos model.
Key Points:
- OSFI (Office of the Superintendent of Financial Institutions) warned banks about the impact of advanced AI models like Anthropic's Claude Mythos.
- The email, obtained by Reuters, noted that these models significantly compress the time frame for risk mitigation, challenging traditional patching cycles.
- This timing aligns with recent discussions between Canadian bank executives and regulators about Mythos, as well as similar conversations at other central banks worldwide.
- OSFI emphasizes it doesn't regulate AI models directly but focuses on how financial institutions manage risks associated with their use.
Access to the Model:
Claude Mythos is a restricted access model distributed through Anthropic’s Project Glasswing. While the Canadian government has access, the usage by individual banks remains unclear.