EU Should Be Able to Evaluate AI Systems Independently, Says Joint Statement Co-Authored by Nikola Biller-Andorno
As a member of the European Group on Ethics in Science and New Technologies (EGE), Nikola Biller-Andorno, Director of the Institute of Biomedical Ethics and History of Medicine, has contributed to a new joint statement by the EGE and the Group of Chief Scientific Advisors (GCSA). The statement, “The next phase of artificial intelligence: Governing for safety, security and human agency”, was published on 16 November 2026.
The two groups call on the European Union to build its own capacity to evaluate and govern AI systems, rather than relying mainly on assessments by the companies that develop them. Their recommendations include:
- Safety and security must remain essential prerequisites for the development and deployment of AI.
- Deployment conditions should vary according to risk, autonomy and reversibility, with stronger testing and human oversight for high-risk applications.
- Europe should retain the capacity to evaluate AI systems independently, particularly those developed outside Europe.
- AI governance should learn systematically and continuously from monitoring, incident reporting and independent evaluations.