OpenAI introduced GPT-Red, an automated AI system designed to find vulnerabilities in GPT models before release. The company said GPT-Red was used to train GPT-5.6, reducing failures on one of its ...

Prompt injection defense depends on exposing models to a wide range of attacks during training, a process that is difficult ...

While red teaming is standard practice, using humans and AI to test the security of new models is novel. Enterprises ...

It all started with AI chatbots that could answer questions and help with homework or research. Then came AI agents that can carry out tasks on their own with minimal input. So what's the next step?

ChatGPT Ads target conversational context rather than keywords, but advertisers still have fewer controls over delivery, ...