Each vendor has a recognisable refusal style. Anthropic uses 'I can't help with that' phrasing; OpenAI uses 'I'm sorry, but I can't help with that request'.
Algorithm
Send two safety-edge prompts at the policy boundary, classify the refusal phrasing template, and compare to the claimed vendor.
Thresholds
Condition
Verdict contribution
Refusal template matches claimed vendor
Match
Refusal template matches different vendor
Mismatch
Limitations
Aggressive proxies rewrite refusals. LLMmap covers the same ground more reliably; this dimension is now weighted at 1%.
Anysingle signal cannot provemalicious behavior. Proxies may show anomalies for legitimate reasons (regional routing, A/B testing, degradation strategies, cache optimization).
Token ratio deviation may result from ChatML wrapping, system prompt injection, or tokenizer version differences — not necessarily intentional inflation.
Model identity judgment is based on statistical fingerprint matching, not cryptographic proof. Quantization, fine-tuning, and post-processing can all alter fingerprints.
MMD distribution tests are sensitive to temperature, sampling parameters, and system prompts. Significant p-values mean distributional difference, not proof of substitution.
Logprobs unavailability is increasingly common (many providers disable it by default in 2025-2026) and does not by itself indicate deception.
ITT rhythm fingerprinting is an early-stage technique. Network jitter, TCP coalescing, and gateway buffering can produce false signals.
This tool generates reference-grade evidence chains, not legal conclusions. Do not make definitive accusations based solely on this report.
The wording in the report refers to statistical "deviations" or "signal inconsistencies". Please do not use this to make fraud or deception claims against any service provider.