On 4 November, KnowBe4 announced “Agent Risk Manager,” a defensive system for protecting, monitoring and managing the behaviour of autonomous AI agents. It enables risk to be visualised and measured on the KnowBe4 platform, with the company aiming to begin offering it in the Japanese market in fiscal 2027.
On 4 November, KnowBe4 announced “Agent Risk Manager,” a defensive system for protecting, monitoring and managing the behaviour of autonomous AI agents. In response to the growing use of AI agents to view inboxes, access confidential data and execute tasks autonomously, the system enables their behaviour to be visualised, analysed and assessed, while detecting and blocking activity in real time.
Discover provides comprehensive detection and inventory management of shadow AI and AI-agent use within an organisation through browser extensions and API integration with major AI platforms, and prevents deactivation without requiring software to be installed on endpoints. Govern applies security policies based on detected risks and controls AI access destinations through permission, warning and blocking functions. Coach provides employees whose risky actions have been detected with real-time notifications and explanations of the reasons, encouraging safe use.
Key features include integration into a common risk score, real-time threat protection and confidential-information masking, feedback on risk details through Slack/Teams/Google Chat, detection and automatic masking of the 4 major AI threats under OWASP standards, discouraging the use of personal accounts and directing users to safe alternative tools, direct integration with major AI platforms, and visibility from the first day of deployment through browser extensions and API integration.
The product is intended for global rollout, with the company aiming to begin offering it in the Japanese market in fiscal 2027.