Skip to content

Release Notes

Latest release

Version 9.2.25.1

Release date: 1 June 2026

This release adds Runtime on Kubernetes as a production deployment option for the AI Assets Collector, alongside the existing Docker path for quick tests and local evaluation.

AI Governance

  • Runtime on K8s — Deploy the AI Assets Collector through a Trust3 Runtime in your own Kubernetes cluster (for example AKS). The Runtime Plane maintains an outbound-only connection to Trust3; collectors are added from the dashboard and run as pods in your cluster.
  • Partially Managed (D2P) runtime — Create a runtime with Basic Config, AWS or Azure as the cloud provider, and install via the published curl/wget and Helm steps until Connection Established is shown.
  • Collector configuration in the portal — Configure Databricks and/or Azure datasource values (and Kubernetes settings) when adding AI Assets Collector from Settings > Runtime > Collectors.
  • AI Assets Collector (Run Collector) — Documentation now covers Docker (docker-compose.yml, .env) and Runtime on K8s in a single Run Collector guide; quick-start and tutorial flows point to the path that fits your environment.

Version 9.2.17.2

Release date: 11 April 2026

This release centers on Trust3 AI Governance: visibility and control over AI agent data access across your estate—discovery, identity mapping, unified inventory, policy-oriented workflows, and natural-language GIA exploration over the same data the dashboard uses.

AI Governance

  • Multi-platform agent discovery — Inventory AI-related assets from connected platforms (including Databricks serving endpoints and Genie spaces, and Microsoft Azure / Power Platform for Copilot Studio and related signals, depending on configuration) with a path for custom registrations where automation is not yet available.
  • Identity discovery and mapping — Associate agents with owners and invokers from platform metadata; support ephemeral identities with parent linkage for a clear chain of custody in reviews and exports.
  • Unified dashboard — Single operational view of agent inventory across platforms, with identity columns and policy posture so IT, platform engineering, and data governance align on the same inventory.
  • GIA (governance Q&A) — Ask plain-language questions over inventory and identities (for example, agents without owners, or agents and their identities) without switching between vendor admin consoles.
  • AI Assets Collector (Run Collector) — Docker-based collector to ingest inventory into AI Governance using docker-compose.yml, .env configuration, and AI_GOVERNANCE_API_KEY where your deployment requires it; documented flows for Databricks and Azure prerequisites and credentials.

Documentation and experience

  • Release notes and primary getting-started material are aligned to the AI Governance journey (inventory, collector, prerequisites) rather than legacy IQ-only flows.

Reporting issues

If you encounter any issues or bugs, please report them through the support portal or contact the Trust3 team at support@trust3ai.com.

Next steps