How to Stop AI Data Leaks: A Webinar Guide to Auditing Modern Agentic Workflows

**The Hacker News** piece highlights that autonomous AI agents — described as 'invisible employees' with broad access — create a new attack surface that attackers can manipulate to exfiltrate data or perform actions. It promotes a webinar (‘Beyond the Model: The Expanded Attack Surface of AI Agents’) that will explain how agents can be abused and provide guidance on discovering and securing them.