Can we use ElevenLabs for SPYN's voice features?

4 messages · fast mode

  1. Can we use ElevenLabs for SPYN's voice features?

  2. Yes, with conditions. Three skills apply: GDPR (voice is personal data, biometric if cloned from a sample), EU AI Act (synthetic audio under Article 50), and Architecture Decisions (this needs an ADR).

    legal/gdprlegal/eu-ai-actdeveloper/architecture-decisions

  3. What's the ADR look like?

  4. I drafted one. Title: 'Adopt ElevenLabs as runtime voice provider'. Posture: ADR-required because we add a new sub-processor that produces synthetic content. Want me to open the PR?

Prototype

Reply input is mocked — production lets you continue the thread.