Trust3 AI adds agent security layer for Microsoft Fabric and OneLake

Trust3 AI said Oct. 2 it has added a capability that syncs row- and column-level security across Databricks, Snowflake and Microsoft OneLake Security. The move targets a major blocker for enterprise AI: keeping autonomous agents governed well enough to move from pilot to production in Microsoft Fabric environments.

Why it matters: - Enterprise AI programs can stall when autonomous agents cannot inherit consistent access controls across data platforms. - Trust3 AI is aiming to reduce that risk by extending governed access into Microsoft Fabric and OneLake. - The company says the capability is meant to help enterprises move AI agents from pilot to production at scale.

What happened: - Trust3 AI announced a new capability on Oct. 2, 2026. - The capability automatically synchronizes row- and column-level security policies across Databricks, Snowflake and Microsoft OneLake Security. - The announcement came during FabCon and SQLCon Barcelona 2026. - Microsoft also highlighted an expansion of the OneLake ecosystem at the event.

The details: - Trust3 AI said the system carries authorization intent from Databricks and Snowflake into Microsoft Fabric. - The company said that creates one consistent security posture across access paths, including access by autonomous AI agents acting on a user's behalf. - Trust3 AI said the approach avoids rebuilding access policies separately in each environment. - Don Bosco Durai, CTO of Trust3 AI, said enterprise AI agents should inherit the same trust boundaries as the people and systems they support. - Durai said the Microsoft OneLake Security integration gives enterprises a way to scale autonomous AI without building a parallel security model or weakening existing controls. - Microsoft announced multiple Fabric and OneLake updates at FabCon and SQLCon Barcelona 2026. - Those updates included IQ sharing in preview, bidirectional Salesforce Data 360 integration, lakeFS versioned data access, Dynamics 365 Business Central mirroring, production-ready Azure Databricks native storage in OneLake, Google BigQuery mirroring, on-demand OneLake compute billing, Fabric zero-provisioned (F0), a OneLake Table Read API for AI agents, public preview of ClickHouse workloads in Fabric, deeper Microsoft Foundry and Excel integration, and new partner workloads from Esri, Telmai, Celonis, Spectral Core, Financial Fabric, IntuigenceAI and Metropolis. - Microsoft’s Dipti Borkar said access controls cannot stop at platform boundaries as enterprises adopt a best-of-breed data strategy. - Borkar said consistent security across OneLake, Databricks, Snowflake and every access path is essential for people and AI agents to work with data without losing governance. - Trust3 AI said its platform provides a unified control plane for AI agents and data, with governed access, agent discovery, real-time observability and runtime security. - Trust3 AI said the platform is designed to help enterprises scale autonomous AI without duplicating access policies.

Between the lines: - The announcement reflects a larger shift in enterprise AI from model capability to operational control. - Security and governance are becoming gating factors for deployment, not after-the-fact add-ons. - Microsoft’s broader OneLake push suggests the company is positioning Fabric as a governed hub for both human users and AI agents. - Early partner onboarding to IQ sharing shows Microsoft is trying to make governed data sharing more usable across ecosystems. - TomTom and EY are among the first partners onboarding to IQ sharing. - TomTom Chief Product Officer Leo Sei said trusted location data is a foundational ingredient for customer applications, AI solutions and intelligent agents. - Sei said TomTom is excited to extend access to its location data products across customer analytics and AI ecosystems. - EY Vice President of Cloud Marketplace John Whittaker said IQ sharing could make context reusable, governed and portable. - Whittaker said the approach could bring trusted knowledge, methods and domain insight to clients in new AI-enabled ways.

What's next: - Trust3 AI is directing readers to its website, technical deep dive and demo request page for more information. - The company is positioning the new capability for enterprises that want to scale autonomous AI inside Microsoft Fabric without loosening governance. - Microsoft’s expanding OneLake ecosystem suggests more partner integrations and governed data-sharing features are likely to follow. - Trust3 AI said its platform is available at More information, with a technical deep dive and demo request page.

The bottom line: - Enterprise AI may not reach production at scale unless security follows the agent everywhere the data goes.