Skip to main content

EU AI Act

This page explains how Sally AI complies with the EU AI Act (Regulation (EU) 2024/1689), our risk classification, and the transparency obligations we fulfil under Art. 50. Use it for vendor assessments, compliance reviews, or to verify our AI-specific obligations. The full compliance declaration is available in the Download Center.

Risk classification
Limited Risk
Art. 50 transparency obligations apply

Under Art. 50 of the EU AI Act, Sally AI is classified as a Limited Risk system: it interacts directly with natural persons, so it must clearly disclose that it is AI. Sally AI meets all applicable Art. 50 transparency requirements, see below.

Sally AI is not a high-risk system under Annex III and does not use any prohibited practices under Art. 5. In particular, no emotion recognition in the workplace, no biometric categorisation, and no remote biometric identification.

How we fulfil Art. 50 transparency obligations​

✓After joining a meeting, Sally posts a transparency notice in the chat that clearly identifies it as an AI system and states that transcription is in progress.
✓The notice links directly to the privacy information so every participant can review it in one click.
✓An immediately effective right to object is available: any participant can type opt out in the meeting chat, all captured data is deleted and Sally leaves the meeting.

Prohibited practices under Art. 5, not used by Sally​

Sally does not implement, and is not marketed for, any of the practices prohibited under Art. 5 of Regulation (EU) 2024/1689. Confirmed exclusions:

  • No emotion recognition in the workplace or educational settings.
  • No biometric categorisation to infer race, political opinions, trade union membership, religious or philosophical beliefs, sex life or sexual orientation.
  • No remote biometric identification. Sally does not identify individuals from face, voice or other biometrics.
  • No social scoring, no manipulative subliminal techniques, no exploitation of vulnerabilities.
  • No use in high-risk decision-making (employment, credit, law enforcement, migration, education, essential services).

The full AI compliance statement is available as Annex 5 of the DPA (download). Detailed obligations under Regulation (EU) 2024/1689 are covered in Annex 6 of the DPA (download).