EU Lawmakers Vote to Ban Public Face-Scanning, Stricter AI Regulations

The European Parliament has passed a crucial resolution that outright bans real-time, remote biometric monitoring. Due to this decision, future discussions might put legislators in conflict with the EU's 27 member nations.

Last week, the failure of a political agreement cast doubt on the broad prohibition, which MPs had previously approved. Lawmakers also established new regulations about general-purpose AI and pillar models like GPT-4 in addition to the restriction on biometric monitoring.

The parliament's proposal calls on businesses like OpenAI Inc. and Google, regardless of the usage of their AI models, to do risk analyses and disclose summaries of the copyrighted materials they used to train such models, according to Bloomberg.

European Parliament President Roberta Metsola stressed the need to define boundaries and restrictions for artificial intelligence at a news conference. She said technological breakthroughs must always be consistent with democratic principles and fundamental rights.

What will happen next?

The parliament approved the whole text of the AI Act with 499 votes in favor, 28 votes against, and 93 abstentions. This opens the way for later-scheduled "trilogue" deliberations between the parliament, EU member states, and the European Commission.

Any product or service using AI systems would be subject to regulation under the AI Act, which was first proposed in 2021. It divides AI systems into categories depending on the degree of danger, which ranges from minimum to intolerable. Transparency and precise data use will be required for high-risk applications like employing or targeting children.

The EU member states will enforce the regulations, and authorities can compel businesses to take their AI-based goods off the market. Fines for breaking the rules may exceed 30 million euros ($33 million) or 6% of a company's yearly worldwide sales, amounting to billions for digital giants like Google and Microsoft.

Protecting The Public From AI Threats, Upholding Democracy

The European Union's primary goal for the move is to protect people's lives, property, and fundamental rights and values against AI-based harm. Such "social scoring" systems, which rank people according to their actions, are among the applications of AI that are explicitly forbidden. Exploitative AI that targets vulnerable populations, such as children, or uses subliminal manipulation to promote risky conduct is likewise prohibited, per Time.

The compromise version of the AI Act permits remote biometrics to be used by law enforcement, but only after a significant crime has been committed, with court approval, and not in real-time.

Several Members of the European Parliament (MEPs) believe that using biometrics by law enforcement is incompatible with democratic norms, which has made this issue divisive. The topic has been driven by China's Communist Party's extensive use of face recognition cameras to monitor the people, according to Euronews

During the voting, President Roberta Metsola emphasized the need to balance technical progress and society's interests by reiterating the necessity for distinct boundaries and restrictions on artificial intelligence.