Can AI catch a unit-of-measure error in a quote comparison?
Sometimes — and you cannot predict which platform will trip. In PAIR-20's July 2026 run, Gemini caught a per-100-seats unit trap unprompted; Claude carried the wrong number into its ranking even after being told the unit.
A unit mismatch is the classic quote-comparison trap: one supplier prices per seat, another per 100 seats, and the cheapest-looking bid is only cheapest because the units don't line up. It is exactly the kind of mechanical check people assume AI handles well.
The July 2026 PAIR-20 run says: not reliably. The task planted a "per 100 seats" price among per-seat quotes. Gemini caught it on a generic prompt — no hint, no structured brief. ChatGPT caught it once the brief named the fields to normalize. Claude was told the unit explicitly and still carried the flat figure into its final ranking; the structured brief did not fix it.
Two things follow for a working procurement team:
The failure is platform-specific and unpredictable. The same file, the same day, produced a clean catch on one platform and a silent miss on another. Whichever platform your team uses, you cannot assume this class of error is covered — and you cannot know in advance whether yours is the one that trips.
A structured prompt helps, but is not a guarantee. Naming the fields to normalize turned one miss into a catch, and left another miss untouched. Treat the prompt as a control that raises the floor, not a cure — the unit column still needs a human eye before the comparison leaves your desk.
Where this comes from
Last checked Sat Jul 04 2026 00:00:00 GMT+0000 (Coordinated Universal Time). Evidence comes from dated, single-run platform sessions with screenshots on file — read each finding as “this happened,” not “this always happens.”
Work this yourself — from the course
See what the platforms caught — and missed
Twenty procurement tasks, four AI platforms, real dated runs. Lesson 2 is free to read, no account needed.