Collibra acquires trail ML to automate AI governance

Collibra announced on Oct. 5, 2026, that it had acquired trail ML, a Munich-based company focused on AI governance. Trail ML was founded in 2023 by Anna Spitznagel, Nikolaus Pinger and Sven Hölzel. The companies present the combination as a way to automate governance work across the AI lifecycle, from policy assessment to enforcement.

Collibra contributes enterprise context and control covering data, models, applications and agents, while trail ML adds agent-powered processes for examining evidence and applicable requirements. Those processes can assess controls, flag gaps and automate related governance workflows, according to the announcement. The stated purpose is to help organizations turn governance policies into ongoing operational steps rather than leave them as requirements to interpret. For teams building AI systems, that points to a governance workflow that can keep checking controls as evidence and agent activity change.

Trail ML agents can examine an AI system's context, identify relevant frameworks and controls, and assess whether those controls are in place and effective. Examples cited include the EU AI Act and ISO 42001, as well as the NIST AI Risk Management Framework. Collibra says assessments can be triggered again when supporting evidence changes, creating a more frequently updated view of governance and compliance than manual, point-in-time reviews. The acquired company's runtime capabilities are designed to enforce Collibra policies where agents operate and block actions that violate them before they happen.

The announcement's rationale is that AI incidents can begin with agents taking actions, rather than models merely producing answers. The release describes governance coverage across data, models, applications and agents, alongside automated control assessment and runtime enforcement. For AI builders, these capabilities address two distinct stages: reviewing whether controls exist and applying policy while an agent acts. Collibra and trail ML say combining enterprise context with automation and runtime capabilities is intended to help organizations scale AI safely.