safeguard

OpenAI’s latest AI models have a new safeguard to prevent biorisks

OpenAI says that it deployed a new system to monitor its latest AI reasoning models, o3 and o4-mini, for prompts…