Press Releases August 3, 2026 09:00 AM

Varonis Introduces Agent Intent-Based Access Control

Varonis launches Agent Intent-Based Access Control to enhance AI security in enterprise data access

By Sofia Navarro
Share
Twitter Reddit Facebook LinkedIn
VRNS

Varonis Systems, Inc. announced a new feature in their Atlas platform called Agent Intent-Based Access Control (IBAC), designed to monitor and control AI agents' access and actions on enterprise data. This innovation aims to prevent unauthorized or risky behavior by AI agents, addressing concerns about data breaches and operational disruptions caused by rogue agents. Agent IBAC evaluates agents' reasoning and actions against their given instructions, enabling organizations to detect intent drift, quarantine suspicious sessions, and maintain a full audit trail for compliance and governance.

Varonis Introduces Agent Intent-Based Access Control
VRNS
Summarize with
ChatGPT Perplexity Claude Grok Gemini

Key Points

  • Introduction of Agent Intent-Based Access Control (IBAC) within Varonis Atlas to secure AI agent interactions with enterprise data.
  • Capabilities include intent drift detection, full session evaluation, quarantine options, human-in-the-loop approvals, and comprehensive audit trails.
  • Compatibility with widely-used AI agents like Claude Code, GitHub Copilot, and Microsoft Copilot Studio enhances broad enterprise applicability.

MIAMI, Aug. 03, 2026 (GLOBE NEWSWIRE) -- Varonis Systems, Inc. (Nasdaq: VRNS), the data and AI security leader, today 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.

Agents need broad access to data and tools to be useful, which is precisely what makes them risky, and role-based access control was never built to judge what a non-human identity does with the access it has.

"The question is no longer ‘Can a user access this data?’ but ‘In this context, should this agent be allowed to take action on this data?’" said Ron Bennatan, VP of AI and Data Security Strategy at Varonis. "We built Agent IBAC to give organizations assurance that their agents are acting as intended and not putting their business at risk."

How Agent IBAC Works

Agent IBAC checks whether an agent’s reasoning, tool use, and data access are consistent with its assigned instructions, and alerts on or blocks actions that do not align.

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

Key Capabilities:

  • Intent drift detection: Compares the instruction an agent received to its reasoning, tool calls, and data access with lenient, balanced, and strict sensitivity settings.
  • Full session evaluation: Reviews every prompt, response, and tool call in a session to catch risk that builds gradually, such as multi-turn jailbreak attempts. Customers can also define their own session policies in plain language.
  • Quarantine: Blocks an identity or session after a violation for a window the customer sets, with admin controls to lift, extend, or make it permanent.
  • Human in the loop: Routes a flagged action to a person for approval rather than blocking it outright.
  • Complete audit trail: Records every prompt, response, and tool execution alongside the action Atlas took, for security, governance, and compliance teams.

Agent IBAC works with the agents and AI tools organizations already run, including Claude Code, Cursor, GitHub Copilot, and Microsoft Copilot Studio. It is available today to Varonis Atlas customers and is part of the platform's broader approach to securing agents from code to runtime.

Additional Resources: 

  • Learn more about Agent IBAC.
  • Request your Varonis Atlas demo. 
  • Visit our blog, and join the conversation on LinkedIn and YouTube. 

About Varonis 
Varonis (Nasdaq: VRNS) secures AI and the data that powers it. The Varonis platform gives organizations automated visibility and control over their critical data wherever it lives and helps ensure safe and trustworthy AI from code to runtime. Backed by 24x7x365 managed detection and response, Varonis gives thousands of organizations worldwide the confidence to adopt AI, reduce data exposure, and stop AI-powered threats. 

Investor Relations Contact: 
Tim Perz 
Varonis Systems, Inc. 
646-640-2112 
[email protected]

News Media Contact: 
Rachel Hunt 
Varonis Systems, Inc. 
877-292-8767 (ext. 1598) 
[email protected]


Risks

  • AI agents may still pose unforeseen security challenges, especially with complex or sophisticated intent drifts that bypass detection.
  • Implementation complexity and potential false positives could impact operational workflows, requiring careful tuning and human oversight.
  • Reliance on AI agent monitoring intersects with data security and compliance sectors, both of which are sensitive to technological and regulatory changes.

More from Press Releases

Home BancShares, Inc. Announces Recognition in Forbes Top-Performing Banks and Bank Director RankingBanking Sep 18, 2026 New Teva Study Identifies the Attributes that Healthcare Providers Prioritize when Selecting a Treatment for Tardive Dyskinesia in Older Patients Sep 18, 2026 Teva Highlights New Real-World Data Showing UZEDY® (risperidone) Outperforms Other Long-Acting Injectables in Keeping Schizophrenia Patients on Treatment, While Reducing Emergency Care and Healthcare Costs Sep 18, 2026 Teva Announces New Olanzapine LAI (TEV-'749) Post Hoc Data Highlighting High Rates of Stabilization and Long-Term Efficacy as Once-Monthly Subcutaneous Injectable for the Treatment of Schizophrenia in Adults Sep 18, 2026 Enact Receives Ratings Upgrade from AM Best Sep 18, 2026