Italy switches on real-time facial recognition as EU calculates risk to basic rights

A Legislative Decree amending the AI Act came into force in Italy this week, establishing harmonized rules on the use of AI systems, including live biometric recognition technology.

A report from 24 (translated from Italian using AI) frames Italy’s approach as human-centric in nature. “Even when AI is used by the police, the guiding principle is that artificial intelligence serves as a tool to support human intervention and decision-making,” it says.

“Particularly for high-risk systems, the regulation requires constant human oversight by personnel trained in the use of the technologies to prevent security risks or infringements of privacy and fundamental rights.”

The third section of the decree is flagged as the most controversial change. It concerns real-time biometric identification; specifically, “artificial intelligence systems used in police work for the tagging, filtering and categorization of biometric data, for remote real-time biometric identification for the purposes of prevention or protection, and for retrospective facial recognition for the purposes of combating crime.”

The new rules limit the “labelling, filtering and categorization of biometric data within precise and stringent limits and solely for police purposes of comparison and investigation.” The use of real-time facial recognition in public spaces is limited to preventing specific threats or to search for missing persons.

Per the report, “it is essential that databases are created specifically for the event, are deleted upon expiry of the statutory time limits, and lack incremental capabilities compared to previous authorizations.” Web scraping for images is also prohibited.

FRA raises concerns about infringement on rights

Italy’s deployment comes amid a flurry of activity over live facial recognition in Europe, where the European Agency for Fundamental Rights (FRA) is monitoring concerns related to privacy and surveillance. It has published a new report, “Remote Biometric Identification for Law Enforcement Purposes: Fundamental Rights Implications.”

Arguing for storing safeguards built into frameworks from the ground up, FRA says its research has shown “how quickly measures framed as exceptional can start to feel routine, particularly when it comes to Member States strengthening their internal security. That is precisely why strict safeguards must be built in from the outset, not added once a system is already in use.”

The report is comprehensive in looking at case studies from France, Germany, Greece, Hungary, Italy and the Netherlands. It lays out 11 detailed opinions on how the EU should proceed, and it is clear in stating of live biometric matching that “this AI-driven power raises serious fundamental rights issues.”

Yet the most precise and accurate summary and takeaway is included as a pull-quote from a data protection authority that participated in the survey.

“You can’t just, in our view, replace old-fashioned police work because it’s convenient with a technology that is highly intrusive and indiscriminate in its application, that doesn’t meet the thresholds of balancing the rights of individuals, on the basis of strict necessity.”

Article Topics

AI Act  |  biometric identification  |  Europe  |  facial recognition  |  Italy  |  law enforcement  |  live facial recognition