AI audit vs rubric — AI Dropship
An independent Workers AI LLM scored AI Dropship against the same published rubric. The deterministic rubric result is our canonical score. The LLM's result is shown here as a sanity check — never mixed into the scoring formula.
| Dimension | Rubric | LLM | Δ (LLM − Rubric) |
|---|---|---|---|
| Pricing transparency | 84 | 20 | -64 |
| Business transparency | 25 | 0 | -25 |
| Shipping clarity | 80 | 40 | -40 |
| Public reviews | 80 | 75 | -5 |
| Product range | 62 | 55 | -7 |
| Access & onboarding | 85 | 60 | -25 |
| Support track record | 75 | 75 | 0 |
| Store integrations | 55 | 40 | -15 |
| Overall | 69 | 48 | -21 |
What this means: Large disagreement — investigate. The LLM read the published signals very differently from the deterministic rules.
Median per-dimension |Δ| = 20.
Business transparency is 0 due to lack of publicly listed information. Access is 60 due to free trial only. Integration is 40 due to only one platform.This is the LLM's own explanation, not editorial commentary from SupplierSpy. The LLM result is a sanity check on the rubric — never mixed into the scoring formula.