This is a joint project with Ana Oprescu (Science Park) and is titled 'SafeGenAI: Safe Generative AI tools for Creative Industries: current and future practices.'

SafeGenAI: protecting creativity in an AI-driven world

 The SafeGenAI project develops new empirical ways to measure how creative generative AI really is. By embedding these novel creativity metrics as traceable identifiers within privacy-enhancing technologies, the researchers deliver a flexible, attribution-aware (designed to recognise, track and record who created what) framework that helps creators protect their work from unauthorised use.

Practical tools for creators

Together with 2 postdoctoral researchers (1 at ABS and 1 at the Faculty of Science), the team will build tools that allow creators to:

• Set clear usage conditions for their work;
• Trace how content moves through AI-enabled pipelines;
• Generate evidence such as attribution records and audit trails to support enforcement and responsible reuse.

This flexible, attribution-aware framework supports both innovation and accountability.

Interdisciplinary team and partners

SafeGenAI brings together an interdisciplinary team including University of Amsterdam researchers Sander Klous, Martin Senftleben, Tom van Engers, Stevan Rudinac and Nachoem Wijnberg. Industry partners include the Netherlands Institute for Sound and Vision, Stichting Privacy First, Cyber Dune, Calango and Storm Post-Productions.