Yesterday, Varonis announced Agent Intent-Based Access Control (IBAC), a new capability in Varonis Atlas that lets businesses connect AI agents to their enterprise data with safeguards that stop dangerous or out-of-policy behavior.

Agents are making headlines for going rogue, exposing sensitive company data and, in one case, deleting an entire production database.

Agent IBAC compares the instruction an agent received to its reasoning and the tools and data it reaches for, then responds in real time to actions that don't align, including alerting or blocking.

When an agent crosses the line, Atlas can quarantine the identity behind it and block everything that follows for a defined window.

Agent IBAC can be tuned to take appropriate action based on the potential impact. For example: