Why industrial buyers rate AI autonomy before accuracy

How much autonomy a buyer will grant can be scored on its own, and it gates the deal before anyone discusses model accuracy.

By Harinderpal Hanspal on March 2026. Updated October 2026

Industrial buyers decide how far an AI system may act before they weigh how accurate it is. Thing Company scores that willingness as its own dimension, autonomy tolerance, from one to five across discovery interviews. A strong signal needs an average of 4.0 or above, the same threshold every other dimension uses.

Autonomy tolerance is scored one to five, and a strong signal starts at 4.0 A one-to-five scale with a red threshold at 4.0. Below it, the viable product keeps a human in the loop; at or above it is a strong signal. Three decisions by one plant manager sit at different points: flagging an anomaly high, rescheduling maintenance in the middle, stopping a line low. 1 2 3 4 5 4.0 Below 4.0: keep a human in the loop The system recommends and the operator confirms; more autonomy is earned against an operating record 4.0 or above strong signal Stop a line Reschedule maintenance Flag an anomaly One plant manager, three decisions. Positions are illustrative. AUTONOMY TOLERANCE, SCORED IN INTERVIEWS
Autonomy tolerance is scored one to five, and a strong signal starts at 4.0

Trust varies by decision. The same plant manager may trust a system to flag an anomaly, hesitate to let it reschedule maintenance, and refuse to let it stop a line. It changes with the use case, the site and the consequence of a wrong action, and most of all with how much autonomy is requested.

A low score is a product-configuration finding, and arguing with it wastes a sales cycle. At that account the viable product may keep a human in the loop: the system recommends, the operator confirms, and more autonomy is earned later against an operating record.

The 4.0 threshold is Thing Company methodology, not an external benchmark. The market data says only that autonomy arrives in steps: Gartner expects semiautonomous agents to orchestrate 10% of key production, quality and maintenance use cases by 2030, up from 2%. A vendor who prices for full autonomy while the buyer sits on the advisory step creates its own stall.

Go deeper: Governing agents in production: what to ask before an agent acts, AI agent autonomy levels: no standard scale exists, so ask for the vendor's, Where to put the human in an AI agent's work: approve the write, not the draft, Human in the loop for AI outbound agents: review the send, and every route to it and how we score agentic AI covers all four dimensions.