EU to consider ban on AI use for citizen surveillance

The European Union intends to ban AI systems used for mass surveillance or ranking social behavior. The draft law, which the European Commission is due to present in the coming weeks, has been obtained by Bloomberg.

The document envisages a complete ban on AI systems that manipulate people’s behaviour, conduct social scoring, or engage in indiscriminate surveillance. Some security-related exemptions may apply.

In addition, the use of remote biometric identification in public places—such as facial recognition—would require authorisation from authorities.

High-risk AI would be those systems that may threaten safety and fundamental rights, as well as the EU’s democratic values. The authors of the regulation propose assessing them for impartiality.

EU member states will have to establish bodies to test, certify and inspect AI systems. Conformity certificates issued by these bodies would be valid for up to five years.

Companies that develop banned AI services, provide false information or refuse to cooperate with authorities may be fined up to 4% of their revenue.

The regulation will apply to firms based both in the EU and abroad.

However, under the draft, the new rules do not apply to AI systems used for military purposes.

The European Commission declined to comment on the proposed rules.

In April 2021, a group of 56 European organisations urged the European Commission to ban the use of facial-recognition technology without any exemptions.

In January, Amnesty International called for a ban on facial recognition in public spaces. The organisation argues that biometric identification of people via street surveillance cameras constitutes a gross intrusion into citizens’ privacy.

Subscribe to ForkLog news on Telegram: ForkLog AI — all the news from the world of AI!