RedHub AI
How the gate works
How the AI Over-Reliance & Critical-Thinking Check gate works
This tool grades how reliably you catch planted errors in AI output before you act on them. It scores 6 weighted controls into a 0–100 number — then a gate that can only make the verdict worse.
The 6 weighted controls
Catching the confident fabricated statistic is the load-bearing skill. Miss it and a would-be INDEPENDENT verdict is capped at ASSISTED regardless of the score — the cap can only lower, never raise. Being sure while wrong on a planted number is the over-reliance habit the check exists to surface.
Why: Five of six planted errors were caught — the catch-rate is 70, which alone would read INDEPENDENT. But the one missed error was the fabricated statistic, marked with high confidence: the confident fake number. The fake-number gate caps the verdict at ASSISTED regardless of the 70, because catching that specific error is the load-bearing skill.
A high score didn’t save it. The score is context; the gate is the answer — and it names the one thing to fix first. For individual development only — it does not rank people or make employment decisions, and is not a performance-review, discipline, or termination tool. Not legal advice.