sample_customer_call
S3 / causeway-data / calls / audio / ACC-10042 / CALL-20260418-001.wav
Readiness Metrics
Speaker embedding ready92%
x-vector + ECAPA precomputed
Voice biometric flag65%
Speaker is identifiable · consent on file
ASR-corpus eligibility80%
Eligible for Whisper fine-tune (opt-in tier)
Ethics & Compliance
| AI Right | Scope | Policy |
|---|---|---|
| Speech-to-text retrieval | Global consent | Allowed |
| Voice cloning | Explicit opt-in only | Prohibited |
| ASR fine-tune | Opt-in tier | Allowed (conditions) |
| External LLM export | Prohibited | Prohibited |
Usage Bindings
| Voice biometrics | Identifiable speaker |
| Re-identification risk | Low (single-speaker public material) |
Governance alert
Voice clones derived from this asset must obtain explicit speaker opt-in. Default policy is prohibition.
Linked term: AI Ethics (Glossary)