Current reliance on interpreters for video calls creates significant supervision complexity and resource demands.

To address this, STeRn is building an AI-assisted monitoring system for controlled, privacy-preserving videotelephony. The system integrates several key technologies: real-time multilingual speech interpretation and transcription, facial recognition for identity verification, and content moderation via symbol and explicit content detection. This design ensures that the AI intervenes exclusively in cases of rule violation, guaranteeing secure, efficient, and legally compliant communication.

The project’s unique strength lies in combining advanced AI with comprehensive legal, ethical, and social-scientific evaluation to create a human rights-aligned solution, in partnership with the Austrian Ministry of Justice, TU Wien, ONG-IT GmbH, and the University of Innsbruck.