Data-agent platform Wren AI advertises full query-level audit logs, discloses no insurance or compliance-certification terms

Wren AI, an open-core platform that lets both people and AI agents query company data in natural language, says its governance layer is "Auditable by design; not as a workflow, but as infrastructure. RLS and CLS enforced at query time, across every human or agent. Role-based access with full activity logs. Provable who saw what, which query produced which number," according to the company's website. Wren AI says it is "Trusted by 50,000+ users worldwide."

A banking and finance case study on the site describes credit, trading and compliance teams querying separate systems through the same governed layer: "Credit queries the loan book, trading the OMS, compliance the monitoring system, all in plain language. Governed answers in seconds, every query auditable." The company's embedded-API product lists "Signed JWTs Row & column-level scoping Full audit logs Per-app tokens" as standard features, and the platform includes what it calls an "Evaluation Framework 07 Benchmark every answer. Fix every failure."

Wren AI's site does not name a third-party compliance certification -- no SOC 2, HIPAA, ISO 27001 or equivalent -- for its own platform, and does not mention insurance, indemnification, or a liability cap. AI Dispatch's index recorded present values for the company's audit-trail and runtime-governance fields based on the claims above, but no public information for compliance certification, insurance/indemnity, or outcome-settlement -- the same three dimensions still undocumented for most vendors the index tracks, including several, like Wren AI, that market audit-trail features directly into banking and healthcare deployments.

No dispute-resolution process or settlement mechanism is described on the site either, leaving a customer whose "governed" query returns a wrong number with no published remedy beyond contacting Wren AI directly.