Microsoft wants enterprises to closely watch the agents that employees use with systems not controlled by AI labs like OpenAI and Anthropic.
The post Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’ appeared first on TechCrunch.
