Continuously hardening ChatGPT Atlas against prompt injection attacks
OpenAI Blog
December 22, 2025
OpenAI is strengthening ChatGPT Atlas against prompt injection attacks using automated red teaming trained with reinforcement learning. This proactive discover-and-patch loop helps identify novel exploits early and harden the browser agent’s defenses as AI becomes more agentic.
Verticals
airesearch
Originally published on OpenAI Blog on 12/22/2025