Zscaler developing AI-powered breach prediction tool

The ThreatLabz threat research and incident response team at Zscaler is working on an AI-based breach progression prediction tool, which also provides impact analysis and future policy recommendation capabilities, SDxCentral reports. Called Security Autopilot, the tool is currently being employed internally but is set to roll out as a product soon, the company says. Zscaler Product Manager Sanjay Kalra says they are able to understand and predict the path of a malware similar to how a physician can predict disease progression based on experience, but in Zscaler's case after gathering together the more than 300 billion records and transactions that the company obtains daily from network endpoints, partner feeds, threat signals, and other sources and inputting them into Zscaler's proprietary AI engines and large language models for analysis.This enables continuous learning of the constantly changing cloud-based policies and logs. Zscaler can then use these results to recommend policy changes or new preventative measures geared toward an individual type of breach.