The European Parliament has voted a sweeping ban of the use of AI that will affect many facets of smart cities, including public security and safety, and intelligence-led policing.

The amendments are aimed at preventing what Members of the European Parliament (MEPs) describe as “intrusive and discriminatory” uses of AI systems.

The Parliament’s vote has also created differentiations between proprietary and open source AI models. Smart city and policing AI models created under open source licenses are allowed greater scope and functionality under the legislation than proprietary technology.

The votes by the European Parliament introduced substantive new clauses into legislation proposed by the European Commission, extending the scope of the legislation. The new clauses passed votes in both the Internal Market Committee and the Civil Liberties Committee of the Parliament.

The vote prohibits the use of real-time facial recognition technology in public spaces, including using CCTV data as part of a real-time system. The ban also affects using CCTV data within facial recognition databases.

Another clause introduced into the European AI Bill is a prohibition on European Union member state police forces using AI technology for criminal profiling or criminal behavioural analysis. AI-based biometric categorisation systems have also been banned under the amendments.

MEPs justified the new measures as protections for citizens’ health, fundamental rights, and protection of the environment.

Providers of technology using AI foundation models would be required to guarantee fundamental rights, health and safety, democracy, human rights, and protection of the environment. They would also be required to register in an EU database.

The amendments passed in the European Parliament will go to a further vote in the European Parliament in June.