Sweet Security Debuts Agentic AI Blocking To Stop Rogue Agents In Real Time

Sweet Security Ltd., an Israeli runtime security startup, has introduced Agentic AI Blocking, a new capability designed to halt artificial intelligence agents mid-action when they attempt unauthorized tasks in production environments. Announced ahead of Black Hat USA 2026 in Las Vegas, the technology extends runtime security from traditional cloud workloads to autonomous agents operating alongside them. The system can instantly terminate unverified tool calls and active agent sessions, preventing the exfiltration of sensitive corporate data, secrets and personally identifiable information. It also detects and shuts down prompt injection attempts before malicious instructions can alter an agent’s intended behavior inside enterprise environments.