AI research tool THEO offers a free re-run for false claims, discloses no insurance, audit-trail or SLA-remedy terms

THEO, an AI agent-memory and research platform at theogrowth.com, tells customers: "Our guarantee: Every source verified and cited. Find a false claim -- we re-run free," according to its website. The company says reports are "delivered in ~6 hours" and that the tool can "connect THEO to ChatGPT and Claude" so users can "pull, compare, and reshape any report from chat" over the Model Context Protocol.

The free re-run is the extent of THEO's published remedy for inaccurate output. The site does not publish a service-level commitment, a credit or refund policy, or an outcome-linked settlement mechanism tied to accuracy -- meaning a re-run is offered at the vendor's discretion rather than through a defined process. THEO also discloses no runtime-governance controls, no audit trail, no compliance certification, no insurance coverage and no indemnification terms, according to the pages AI Dispatch reviewed.

The gap is notable because verified, cited sourcing is THEO's stated product differentiator, yet the mechanism for when that promise fails is limited to a same-vendor, no-cost repeat rather than a documented dispute process, SLA remedy or third-party backstop.

AI Dispatch's index tracks 21 assurance fields -- covering governance, audit trail, compliance, outcome settlement and insurance/indemnity -- across every entity it publishes, each backed by a source URL, retrieval timestamp and verbatim quote from the underlying page.