Microsoft: Treat AI agents as insider threats - CSO Online

Microsoft warns that AI agents pose cybersecurity risks similar to insider threats that hackers or malicious employees pose. Organizations need to develop policies to protect against AI systems being misused as attack vectors.