AI testing agent KushoAI can block customer software releases; company discloses no insurance or liability terms

KushoAI, an AI agent that runs automated software tests, says it functions as an "AI gatekeeper" inside customers' continuous-integration pipelines, with authority to enforce test quality, coverage and release governance before code ships. The company's site states its platform has run "From 7Mn+ test executions across our customer base."

The company says it is "SOC 2 and ISO compliant" and reports "Regular penetration testing by third-party security firms" on its security page, along with "Granular data access policies with audit logs" governing internal access to customer data.

What KushoAI's published pages do not state: any insurance carrier, coverage limit, indemnification term or liability cap tied to the gatekeeper role, nor a settlement or dispute process for a release the agent blocks or approves incorrectly. Regulatory-alignment claims beyond the two named certifications are also absent. All of those fields register as no public information available in AI Dispatch's index, most recently checked Aug. 26.

The gap sits on the dimensions AI Dispatch tracks as runtime governance and insurance/indemnity. KushoAI's own marketing describes real gating authority over a customer's release pipeline — a decision with direct operational consequences if wrong. Its public materials say nothing about who bears the cost when that decision fails, and no insurance or indemnification language appears on the pages reviewed.

The pattern is consistent with what AI Dispatch has found across other agent categories this month: agents increasingly hold real authority — over software releases, customer records or payments — while the companies selling them publish security and compliance claims but stay silent on what happens, financially, if the agent gets it wrong.