Risk management platform provider KnowBe4 has launched Agent Risk Manager, a solution designed to secure, monitor and govern the behaviour of autonomous AI agents.
The solution has been integrated into the KnowBe4 HRM+ platform, and aims to address an emerging security gap as workflows shift from AI assisted to agent managed.
Agent Risk Manager is designed to prevent AI agents from engaging in unauthorised actions such as sharing sensitive information, identifying indirect prompt injections and inventorying agents. Other capabilities include identifying access permissions and tools that AI agents are able to access, stress-testing AI agents against prompt injection and social engineering tactics, and automatically redacting sensitive data before it can reach an audit log.
The solution can also automatically catalogue agents and tools across tenants without manual input, and produces a filterable record of all agent actions and detections for use during incident response and forensic review.
KnowBe4 Chief Product Officer Greg Kras said a paradigm shift in cybersecurity is required for the AI era.
“The industry has spent years securing the human element, but today, AI…
