Keeper Security has introduced KeeperAI, an AI-driven feature for its KeeperPAM platform that detects, classifies, and responds to cyber threats in real time. The solution automates privileged session monitoring and instant risk mitigation, giving organizations faster protection against AI-powered attacks.
SINGAPORE, 29 AUGUST 2025 – As cyberattacks grow faster, more pervasive, and increasingly automated with artificial intelligence, organizations are struggling to defend sensitive systems from insider threats and privilege misuse. In response, Keeper Security, a leading provider of zero-trust and zero-knowledge Privileged Access Management (PAM) software, has launched KeeperAI, an agentic AI feature designed to monitor and secure privileged sessions in real time.
KeeperAI continuously analyzes session metadata, keystroke logs, and command executions to detect anomalies, automatically classifying threats from critical to low risk. The platform enables organizations to configure custom detection rules, instantly terminate high-risk sessions, and integrate seamlessly with cloud or on-premises AI infrastructure.
“Cyber threats are no longer just a question of if, but when,” said Craig Lurey, CTO and Co-founder of Keeper Security. “KeeperAI allows organizations to automatically monitor and shut down risky activity before it escalates, reducing response times from hours to seconds.”
Key capabilities include automated session analysis, customizable threat policies, flexible deployment options, and integration with leading large language model providers such as AWS Bedrock, Google Gemini, Anthropic, and OpenAI. The system supports SSH-based sessions today, with future plans to add RDP, VNC, RBI, and database protocol support.
Designed as a sovereign AI product, KeeperAI ensures all data remains encrypted and fully controlled by the customer while providing seamless SIEM and SOC integration via Keeper’s Advanced Reporting and Alerts Module. The feature is available immediately for all KeeperPAM customers running PAM Gateway version 1.7.0 or higher.
