A roadmap for AI technology in law enforcement
The world is changing rapidly, and artificial intelligence (AI) is central to the transformation. While AI is capable of revolutionising fields such as housing, healthcare, transportation and education, the deployment of AI by law enforcement agencies (LEAs) raises ethical concerns. The EU-funded ALIGNER project brought together European actors concerned with AI and policing to discuss how to strengthen law enforcement and benefit the public.
Advisory boards and workshops
Communication is key to the aims of ALIGNER, and to facilitate interdisciplinary discussions the project established two advisory boards. One board was composed of LEAs, including the Basque, Munich and Swedish police forces, which were among the project partners. The second board included scientific, industrial and ethical experts. Combined, the advisory boards represented more than 60 experts involved in over 30 research projects. The project identified four areas essential to developing a roadmap for addressing AI technology in law enforcement. For each topic, ALIGNER hosted a workshop to facilitate discussion among stakeholders. Two of the workshops addressed how criminals are using emerging AI technologies and how AI can be used by LEAs to counteract this trend. A third workshop focused on legal and ethical concerns of AI deployment in policing, and a final workshop addressed LEA capability enhancement needs and policy recommendations.
Informing the AI Act
The workshops allowed ALIGNER to get input and feedback from target groups in an effective manner. The relevance and immediacy of these dialogues was particularly helpful considering the European Commission’s broad initiative to address the impact of AI on society. ALIGNER coordinated with other projects funded under the H2020-SU-AI-2020 call, particularly popAI and Starlight. All of these projects serve to support the Artificial Intelligence Act (AI Act), a major piece of European legislation. This confluence of legislation and funded projects was an interesting and fruitful challenge for ALIGNER. According to project coordinator Daniel Lückerath: ”We were working on a project whose objective was to give policy and research recommendations on AI, while the EU AI Act was being developed and discussed in parallel.”
Roadmaps and assessments
Major outputs of ALIGNER include publicly accessible deliverables that support LEAs, policy makers and researchers in meeting AI-related objectives. The policy and research roadmap gives an overview of AI technology needs and challenges faced by LEAs. The roadmap provides nine policy recommendations, such as specific training of law enforcement officers to improve awareness of the risks inherent in AI systems. It also identifies 19 areas where additional research is needed, for example, exploring how AI can impact the fairness and transparency of judicial decisions. A noteworthy outcome of the project is the ALIGNER fundamental rights impact assessment (AFRIA). Composed of two complementary templates that help LEAs identify and assess the impact of an AI system on the fundamental rights of individuals, AFRIA helps to ensure compliance with the AI Act. AI compliance with ethical principles was a major focus of ALIGNER. As Lückerath shares: “AI cannot and will not replace human decision-making, but it is an important tool to support the decisions humans make.” With the AI Act and projects like ALIGNER, the EU is preparing for a fast-approaching, unprecedented future.
Keywords
ALIGNER, roadmap, law enforcement, AI technology, AI Act, AFRIA, ethics assessment