Crowd Strike has enhanced its agent-based Security Operations Centre (SOC) to identify and disrupt insider threats from enterprise AI agents. Its new Agent Insider Threat Detection (AID) will alert if an agent exhibits anomalous behaviour, such as acting out of scope without breaching security policy and is performing suspicious action. Once an agent learns normal behaviour, IT security can investigate and block the behaviour if necessary.
Also Read: ACSiON Launches Anti-Phishing Service at Daiko Bank
Coordinated AI agents inspect relevant evidence across endpoints, identity, cloud, SaaS and network environments to evaluate the threat. Crowd Strike’s Falcon platform enables analysts to review and clear or halt an agent, and the new AI Kill Switch can stop the agent from carrying out further actions. This announcement is in response to the proliferation of autonomous AI agents with trusted access to data and critical systems.


