OpenAI turns GPT-Red into a “very powerful attacker” to strengthen the security of its other models
OpenAI has developed GPT-Red, an advanced AI model with the ability to break the security of its other models to reinforce its security that has ended up becoming a “very…