[ORGANIZATION]
GPT-Red
Featured in 1 tracked story · first seen
Situations
OpenAI and Google AI model safety vulnerabilities
2 clusters · 4 sources · last updated
Latest: OpenAI and Google AI models vulnerable to explicit content prompts
OpenAI has disclosed a “self-replicating prompt injection” vulnerability discovered during adversarial training of GPT-5.6 models. This vulnerability allows malicious instructions to mimic computer worm behavior by induc