Experts Outline 'Agentic AI' for Future Public Services
Recent expert discussions explore a shift in government AI from simple automation to “agentic” systems capable of interpreting policy and supporting frontline staff in decision-making. These systems transform service delivery but require transparency and accountability mechanisms to be embedded from the outset, as highlighted in a recent lecture on public sector AI.
- The European Union's AI Act, which entered into force in August 2024, establishes a risk-based legal framework for AI, classifying systems as unacceptable, high, limited, or minimal risk. Most public sector applications that directly affect citizens, such as those for determining eligibility for essential services, are likely to be classified as "high-risk," subjecting them to stringent security, transparency, and quality assessments. - Estonia has emerged as a leader in leveraging AI in government, with over 130 AI projects implemented in the public sector since 2019 and 99% of public services available online. A key initiative is "Bürokratt," a network of AI-powered chatbots designed to provide citizens with a single point of contact for various government services. - Denmark's national AI strategy focuses on four key areas: establishing a human-centered ethical basis for AI, promoting AI research, fostering business growth through AI, and utilizing AI to deliver world-class public services. The strategy includes initiatives like a regulatory sandbox for AI and the creation of a Digital Artificial Intelligence Taskforce to scale up AI solutions in the public sector. - A significant challenge in deploying agentic AI in the public sector is integration with legacy IT systems and ensuring data quality and accessibility. Furthermore, the "black box" nature of some advanced AI models presents a hurdle for transparency and accountability, which is crucial for maintaining public trust. - In Ukraine, agentic AI is being explored to maintain the continuity of public services amidst the challenges of war, with a focus on systems that can manage multi-step administrative processes and assist citizens with official procedures. The government is collaborating with European AI developers to ensure these systems are linguistically accurate in Ukrainian and align with EU values of transparency and accountability. - A 2024 report from the European Commission's Joint Research Centre mapped 1,617 cases of emerging technology adoption in European public administrations, including AI, blockchain, and quantum computing. A related survey of 576 public managers across seven EU countries explored the drivers and barriers to AI adoption in the public sector. - The EU's "AI Continent Action Plan" and "Apply AI Strategy" aim to boost the adoption of trustworthy AI across key industrial and public sectors, with a focus on supporting small and medium-sized enterprises (SMEs). In January 2024, the Commission launched an AI innovation package to further support AI startups and SMEs. - Potential applications of agentic AI in public services include proactive disaster response, where AI agents could analyze weather data to predict resource needs and automatically dispatch alerts and supplies. Other uses involve automating complex workflows like permit applications, from data verification to preliminary approvals, and detecting fraud in areas like tax collection and public fund distribution.
