OpenAI Introduces Advanced Account Security to Safeguard AI Identities

OpenAI announced Advanced Account Security, an opt-in suite for high-risk ChatGPT and Codex accounts that enforces phishing-resistant authentication (passkeys or hardware security keys), disables email/SMS recovery, excludes account conversations from model training, and shortens session durations; the company partnered with Yubico for discounted security keys and will require these protections for Trusted Access for Cyber participants starting June 1, 2026.