Actus AutomatiséesActus TechAI

Continuously hardening ChatGPT Atlas against prompt injection

Par Krigs

22/12/2025

0 0

Continuously hardening ChatGPT Atlas against prompt injection

🕒 Publié le : 22/12/2025 à 18:35
 |  ✍️ Auteur :
 |  📚 Source : OpenAI News

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.