UK Cities Pilot AI for Public Sector Decision-Making
Local authorities in the UK are piloting the use of large language models and predictive analytics for city-scale decisions, including resource planning. A Cambridge University project, "Decision-making with AI in connected places and cities," is studying these implementations to establish best practices for ethical and responsible innovation. The research emphasizes safeguarding transparency and inclusion as AI becomes more embedded in public services.
- The Cambridge University project is specifically investigating how local authorities use AI for decisions on placemaking, land use, mobility, and sustainable water systems. The Department of Engineering leads this interdisciplinary project, collaborating with experts in urban governance, philosophy, and ethics to develop resources for ethical AI implementation. - Several UK councils are already using AI for specific public services. For instance, Northumberland County Council is piloting "FloodAI" to enhance flood resilience, while Transport for West Midlands has used AI to detect near-miss road collisions. In London, the Borough of Hillingdon uses predictive analytics to identify children at risk of criminal exploitation. - The UK government's national AI strategy includes the "AI Playbook" and a refreshed "Data and AI Ethics Framework" to guide the public sector. These resources provide principles for lawful and ethical AI use, emphasizing transparency, accountability, fairness, and the necessity of meaningful human oversight. - Manchester City Council is focusing on resident inclusion by creating a "People's Panel for AI" to ensure public views are incorporated into the development of AI-powered frontline services. This aligns with a broader government focus on building public trust to facilitate the transition to an AI-enabled economy. - Europe's GovTech ecosystem is expanding, with half of all investment deals in 2024 being AI-related. However, AI startup activity is still limited in key regional policy areas like transport and environment. Case studies from other European cities include a chatbot assisting residents in 37 Danish municipalities and an AI system in Verona, Italy, that analyzes traffic patterns to reduce congestion. - To accelerate AI adoption, the UK government is establishing "AI Growth Zones" and a national team of planning experts to support local authorities with complex data centre projects. This initiative aims to address the lack of specialized expertise in some local planning authorities. - A significant challenge for local authorities is the public's trust in their use of data. A survey by the think tank Onward found that only 31% of the public trust their council to use their data ethically. This highlights the importance of transparency and robust governance in the implementation of AI systems. - The London Office of Technology and Innovation (LOTI) runs an "AI Network" with over 200 officers from London boroughs. This network includes a working group that meets monthly to share learnings and discuss common challenges in adopting generative AI.
