The EU is questioning OpenAI and Anthropic over AI agent security incidents, raising questions about the AI Act’s reach into model development and testing.