CyberArk Introduces Comprehensive AI Agent Identity Security
18 April 2025 · Uncategorized ·
Source: · https://infosecu.technews.tw/2025/04/16/cyberark-ai-agents/
With the rapid growth of artificial intelligence (AI), millions of autonomous AI agents are expanding the attack surface in identity security, creating new challenges for organizations. In response to this evolving landscape, cybersecurity firm CyberArk has announced 'CyberArk Secure AI Agents,' a solution designed to help enterprises adopt an “identity-first” cyber strategy specifically tailored for agentic AIs through their existing identity security platform.
Gartner forecasts that by 2028, up to 25% of corporate cybersecurity incidents will be attributable to the misuse of AI agents—potentially stemming from both external attackers and internal malicious actors. This emerging attack surface is driven by a new type of digital identity: agentic AIs characterized by human-like autonomy combined with machine-scale scalability. Managing their privileged access rights, lifecycle operations, and collaborative interactions presents significant hurdles for traditional security measures.
“As millions of autonomous, adaptable, and interactive AI agents gain privileged access to resources and services,” stated CyberArk CEO Matt Cohen, “companies risk substantial exposure if cybersecurity fails to keep pace with innovation.” He emphasized that basic identity and access control alone will not suffice; agentic AIs must be protected from the outset by integrating human-centric security principles with the scalability and automation required for machine identities.
CyberArk Secure AI Agents offer built-in functionalities designed to support an evolving “agentic workforce,” including:
* Identity management
* Access control
* Lifecycle operations
Furthermore, CyberArk has introduced a new open-source safety toolkit called "CyberArk Labs AI Agent Tool Set" aimed at the developer community. Available on their GitHub page, this toolset enables developers to understand how agentic AIs communicate and identify potential risks while providing real-time certificate configuration capabilities for enhanced development efficiency and security.
Gartner forecasts that by 2028, up to 25% of corporate cybersecurity incidents will be attributable to the misuse of AI agents—potentially stemming from both external attackers and internal malicious actors. This emerging attack surface is driven by a new type of digital identity: agentic AIs characterized by human-like autonomy combined with machine-scale scalability. Managing their privileged access rights, lifecycle operations, and collaborative interactions presents significant hurdles for traditional security measures.
“As millions of autonomous, adaptable, and interactive AI agents gain privileged access to resources and services,” stated CyberArk CEO Matt Cohen, “companies risk substantial exposure if cybersecurity fails to keep pace with innovation.” He emphasized that basic identity and access control alone will not suffice; agentic AIs must be protected from the outset by integrating human-centric security principles with the scalability and automation required for machine identities.
CyberArk Secure AI Agents offer built-in functionalities designed to support an evolving “agentic workforce,” including:
* Identity management
* Access control
* Lifecycle operations
Furthermore, CyberArk has introduced a new open-source safety toolkit called "CyberArk Labs AI Agent Tool Set" aimed at the developer community. Available on their GitHub page, this toolset enables developers to understand how agentic AIs communicate and identify potential risks while providing real-time certificate configuration capabilities for enhanced development efficiency and security.