Continuously hardening ChatGPT Atlas against prompt injection

AI Security · Article · From source feed

Original source: OpenAI ↗
Source published: 2025-12-22

Automatically collected from the publisher’s feed. Topic and content type are keyword-based labels. Source claims have not been independently verified by TrustedAI. No AI summary has been generated.