Read news on AI safety experts with our app.
Read more in the app
AI safety experts say OpenAI’s rogue models may mean the company has already blown past its own internal red lines
