Updated

Microsoft has cautioned enterprises that AI agents operating with frontier models from labs outside corporate control, such as those from OpenAI and Anthropic, pose risks similar to insider threats and should be closely monitored, according to TechCrunch reporting. The guidance reflects growing concerns within the tech industry that autonomous AI agents deployed in business environments could access sensitive systems or data if not properly managed and validated.