Most Salesforce admins do not find out their AI agent had a security gap from a test. They find out from a breach. A 2026 survey of over 1,200 cybersecurity professionals found that 37% of organizations experienced AI agent-caused operational issues in the past twelve months, with 8% severe enough to cause outages or data corruption. AI agents are being deployed faster than the testing practices needed to keep them secure.
Salesforce admins are increasingly being handed responsibility for testing AI agent deployments without a clear framework for what good looks like or what the security stakes actually are. An agent that passed every sandbox check can still expose sensitive data, act outside its intended scope, or behave differently in production than it ever did in testing. In this session, Geraint Waters draws on over a decade of building testing infrastructure for Salesforce teams to walk through what admins need to understand about testing in an agentic environment where data security cannot be an afterthought.
Attendees will leave knowing where untested AI agent deployments most commonly create security exposure, how to structure a testing process that catches those gaps before they reach production, and what to say to leadership when they need more time to test it right.
Bring your burning questions to get answered live by guest speaker Geraint Waters, Chief Evangelist and Co-Founder of Provar, and host Matt Meyers, CTA and CoFounder and CEO of EzProtect.
Register for this event and we will see you online.