OpenAI Uses AI Red Team to Strengthen GPT-5.6 Against Prompt Injection Attacks
Decrypt·
60-second summary
OpenAI is using its AI red team, GPT-Red, to strengthen GPT-5.6 against prompt injection attacks, uncovering vulnerabilities that were exploited to make the model more resistant. GPT-Red's automated testing revealed weaknesses that were addressed, enhancing GPT-5.6's security.
OpenAI said its new automated red-teaming model, GPT-Red, uncovered vulnerabilities that were used to make GPT-5.6 more resistant to prompt injection attacks.