Generated text may repeat known copyrighted material. Which check helps?
Why D is the answer
Protected material detection
Protected material text detection is part of Azure AI Content Safety.
In detail
- Protected material detection flags known copyrighted content.
- Log what is flagged, and review it.
Use case
An illustrative scenario, not a real incident.
A writing app must avoid reproducing song lyrics. Protected material detection flags known copyrighted text.
See it in different ways
mindmap
root((Language tasks))
PII
Azure Language
Translation
Azure Translator
Custom format
Language model with schema
Safety
Groundedness detection
Protected materialflowchart LR S0["Generated text"] --> S1["Protected material detection"] --> S2["Allow or block"]
Key point: Protected material detection flags known text.
Why the other options are wrong
- A. A larger TPM quota
- Quota does not inspect content.
- B. A shorter system prompt
- A shorter prompt does not check outputs.
- C. A Data Zone deployment of the same language model
- A deployment type does not inspect content.