According to Decrypt, executives at OpenAI and Anthropic are conducting internal war‑gaming exercises that simulate the political fallout of a major AI‑driven cyberattack. The simulations aim to outline how the companies would respond if such an event occurred. Participants consider possible regulatory, legislative and public‑relations consequences.
The drills include planning for rapid briefings to members of Congress should the need arise. No official announcement or external coordination is mentioned in the exercises; they remain internal and low‑profile.
