AI reliability vendor NOFireAI caps its own liability at $100 while selling runtime governance and audit-trail assurance

NOFireAI, a vendor selling runtime governance, audit logging and compliance tooling for AI agents running in production, caps its total liability to customers at $100, according to the company's published terms of service. The clause states that the total liability of "the NOFire AI parties" for "all damages, losses, and causes of action" tied to the service "will not... exceed, in the aggregate, $100.00."

The company markets itself as "AI SRE: The Context & Control Model for Production" and sells the product on the strength of controls meant to constrain autonomous agents operating inside a customer's infrastructure. Its site says the platform will "gate every action at runtime," checking "every autonomous agent action against your policy at runtime," and that "every agent action" is "signed and replayable." A separate security page states its audit logs are "immutable, tamper-evident" and "streamed to your SIEM as signed receipts." The company lists its compliance posture as "SOC 2 Type II · GDPR · ISO 27001 in progress" and cites an internal benchmark claiming "89% root-cause accuracy on the AI SRE Benchmark."

None of NOFireAI's public pages disclose an insurance policy, a customer indemnity, or any coverage limit distinct from the $100 cap — the fields AI Dispatch's index tracks for insurance availability and indemnification both return no public information. Disputes are directed to binding arbitration in Wilmington, Delaware, under JAMS rules, set out in the same document that caps liability.

An AI Dispatch index review scored NOFireAI 2 out of 3 on both audit trail and compliance based on the disclosures above, and 2 out of 3 on runtime governance for the policy-based action gating described on its site. Under the index's methodology, those automated scores remain capped at 2 pending human review. No score has yet been recorded for insurance/indemnity or outcome-based settlement.