biorisks

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

OpenAI says it has used a new system to check its latest AI-reasoning models, O3 and O4-Mini, for instructions with…