MEPs Demand Ban on Biometric Surveillance in EU AI Rules

MEPs Demand Stronger AI Regulation, Including Ban on Facial Recognition

A collective of 40 Members of the European Parliament (MEPs), representing various political groups, has formally requested the European Commission to reinforce an upcoming legislative proposal concerning artificial intelligence. Their key demand is a complete prohibition on the deployment of facial recognition and other biometric surveillance technologies within public spaces.

Outlawing Sensitive Characteristic Recognition

Furthermore, the MEPs are advocating for a legal ban on the automated identification of individuals based on sensitive attributes. These include characteristics like gender, sexual orientation, racial or ethnic origin, health status, and disability. They express concern that the utilization of AI in these contexts presents substantial risks to fundamental rights and could exacerbate discriminatory practices.

The Commission is anticipated to unveil its proposed framework for regulating ‘high-risk’ AI applications in the coming week. A draft of this proposal, recently leaked, does not incorporate a ban on facial recognition or comparable biometric remote identification systems in public areas, despite acknowledging widespread public apprehension regarding these technologies.

Concerns Over Mass Surveillance

“The use of biometric mass surveillance technology in public areas is facing widespread criticism due to its potential for misidentifying innocent individuals, systematically discriminating against marginalized communities, and stifling a free and open society. A prohibition is therefore essential,” the MEPs assert in a publicly released letter addressed to the Commission.

They further caution against the dangers of discrimination stemming from the automated deduction of sensitive personal characteristics. This includes applications such as predictive policing and the pervasive monitoring of populations through biometric data.

Potential for Harm and Discrimination

“Such practices can infringe upon rights to privacy and data protection, suppress freedom of expression, hinder the exposure of corruption, and diminish individual autonomy, dignity, and self-expression,” the MEPs emphasize. “These harms disproportionately affect LGBTQI+ communities, people of color, and other groups already facing discrimination.” They urge the Commission to revise the AI proposal to explicitly outlaw these practices, safeguarding the rights of all EU citizens.

The MEPs believe the AI proposal presents a valuable opportunity to specifically prohibit the automated recognition of protected characteristics like gender, sexuality, race, ethnicity, and disability.

Weaknesses in the Draft Proposal

The leaked draft does address indiscriminate mass surveillance, proposing a prohibition on this practice and the implementation of general purpose social credit scoring systems. However, the MEPs contend that lawmakers must go further, identifying weaknesses in the draft’s wording.

They suggest modifications to ensure the proposed ban encompasses “all untargeted and indiscriminate mass surveillance, regardless of the number of individuals exposed to the system.”

Exemption for Public Authorities

The MEPs also voice alarm over an exemption within the proposal that would allow public authorities – and even private entities acting on their behalf – to utilize mass surveillance “in order to safeguard public security.”

“This exemption risks contradicting existing EU legislation and interpretations by the bloc’s highest court regarding data protection,” they warn. “Public security is frequently cited as justification for mass surveillance, and courts have consistently invalidated legislation permitting indiscriminate bulk processing of personal data, such as the Data Retention Directive. This carve-out must be removed.”

They further argue that the proposed language could be interpreted as undermining other secondary legislation previously interpreted by the Court of Justice to prohibit mass surveillance. The MEPs insist the AI regulation must clearly state that its requirements are supplementary to existing data protection laws, not a replacement for them.

Commission Response Anticipated

The European Commission has been contacted for comment regarding the MEPs’ concerns, but is unlikely to respond before the official release of the draft AI regulation, expected next week.

Whether the AI proposal will undergo substantial revisions remains to be seen. However, the MEPs have issued a clear warning that fundamental rights will be central to the legislative debate. They assert that claims of a framework for ‘trustworthy’ AI will lack credibility if the rules fail to address unethical technologies directly.