Europol Report Backs Balanced Approach to AI and Facial Recognition in Policing

A new report from the European Union Agency for Law Enforcement Cooperation (“Europol”) highlights the transformative potential of artificial intelligence and biometrics in policing, while also cautioning about ethical and regulatory risks. Europol is the European Union’s overarching law enforcement agency, so its words carry some weight.

The “AI and policing” report explains that “biometric identification technologies, particularly facial recognition, play a crucial role in law enforcement for prompt and efficient identification of unknown persons.” It argues that the technology has proven particularly useful in criminal investigations, such as solving cold cases or locating missing persons, by comparing images with national or international databases.

A major advantage of facial recognition, as the report describes, is its potential for linking multiple investigations or connecting suspects to previous crimes. For example, Europol’s systems have been instrumental in identifying offenders across multiple cases, particularly in areas like child exploitation.

The report notes: “police confiscate a child sex offender’s computer, initiating a biometric analysis of extracted images. Matches with victims from previous investigations help unveil a broader network of criminals involved in child exploitation.” This demonstrates how biometric systems, particularly facial recognition, can be pivotal in uncovering large criminal networks and securing convictions.

Despite its benefits, the report stresses several ethical concerns associated with facial recognition technology. Privacy concerns are at the forefront, particularly regarding its use in public spaces, where the technology could lead to mass surveillance. The report addresses this issue directly, noting: “the rise of facial recognition has also been followed by concerns. Notably, bias remains a topic of debate… the system’s efficiency, particularly when identifying individuals from specific ethnic backgrounds, genders, or age groups.”

These biases, if not properly mitigated, could lead to wrongful arrests and reinforce existing societal inequalities. This is especially problematic when facial recognition is used in real-time in public settings.

The EU AI Act plays a central role in regulating the use of such high-risk AI systems. The report highlights that “compliance with the EU AI Act represents a crucial balancing act, as it requires law enforcement to adhere to stringent ethical, legal, and privacy standards.” The report suggests that while the technology can enhance public safety, it is essential that law enforcement agencies adopt it in a way that aligns with privacy laws and ethical guidelines. Real-time facial recognition, in particular, must be subject to strict oversight, while retrospective applications—such as reviewing footage after a crime has occurred—are considered more acceptable under well-regulated conditions.

All told, the report calls for a balanced approach. It emphasizes the need for “accountability, transparency, and explainability” in AI-driven policing tools. Law enforcement agencies must ensure that these technologies are used responsibly, within a framework that protects human rights and fosters public trust, as the report states: “policing must navigate the challenges and opportunities of the digital age… biometrics stand out as invaluable allies when used responsibly.”

Source: Europol

–

September 24, 2024 – by Cass Kennedy and Alex Perala