Jerusalem, 17 November, 2026 (TPS-IL) — Israeli-founded cybersecurity company Cato Networks has unveiled AgentSense, a patented technology designed to prevent artificial intelligence agents from taking harmful actions while operating autonomously. Developed by Cato AI Labs, the system monitors agents in real time, detecting malicious instructions hidden in emails, documents and websites before they can influence an agent’s actions. It also aims to identify when an agent’s actions drift from its original task.
Cato, which was co-founded by Israeli entrepreneur Shlomo Kramer, also launched a challenge inviting cybersecurity researchers to attempt to bypass AgentSense, offering $10,000 to each of the first three successful participants.