AI detection helps surface patterns humans miss — but it also fails in edge cases.
That’s why CanonProof treats AI outputs as advisory indicators:
- AI can support suspicion, but not decide alone
- Outcomes are driven by explainable forensic indicators
- When signals conflict, the system can return INDETERMINATE
This approach keeps the result defensible in professional workflows.