AI 'Agents' Arrive for Pro Services

Agentic AI systems represent a significant shift from traditional automation, which follows predefined rules, to autonomous systems that can reason, plan, and adapt to achieve goals. These AI "agents" can manage complex, multi-step tasks, coordinate across different systems, and involve human oversight only when necessary, fundamentally changing how workflows are executed. Rocketlane's Nitro platform, for instance, uses this agentic architecture to not just track, but actively execute delivery tasks like system configuration, data migration, and documentation. In the public sector, this technology offers a path to move beyond the current exploratory phase of AI adoption, where many pilots fail to scale. While a 2025 collection of use cases showed 245 AI initiatives across 21 European countries, many are chatbots or internal automation rather than complex workflow execution. Agentic systems could be applied to grant management, for example, by having an AI agent perform the initial review of reimbursement claims for compliance, flagging exceptions for human review. The European Union's AI Act, the world's first comprehensive legal framework for AI, will heavily influence such applications. The Act classifies AI systems by risk, with applications that assess eligibility for public benefits or assist in legal interpretation categorized as "high-risk," requiring rigorous assessment and transparency. This regulatory framework requires that public sector bodies document their use of AI, ensure human oversight, and manage risks related to bias and opaque decision-making. For UX designers in government, the rise of agentic AI necessitates a focus on service design that accounts for both automated and human touchpoints in a complex journey. AI can accelerate the research and drafting process of customer journey maps by analyzing vast datasets to identify patterns and friction points. However, the designer's role shifts to ensuring these AI-driven workflows are transparent, accessible, and aligned with strategic goals, moving from guesswork to a more data-driven understanding of user needs. This technology also has profound implications for digital accessibility. AI can automate accessibility checks, generate real-time captions, and create adaptive interfaces that adjust to a user's context, such as simplifying navigation for someone showing signs of cognitive load. As public services must be inclusive by design, leveraging AI to meet Web Content Accessibility Guidelines (WCAG) and the principles of the European Accessibility Act is a critical application. The challenge lies in preventing AI models from perpetuating biases, which requires inclusive design practices and diverse training data. Despite the potential, significant barriers to AI adoption in the public sector remain, including legacy IT systems, siloed data, and a shortage of specialized skills. Successful implementation requires more than just technology; it demands a focus on data modernization, clear governance, and upskilling the workforce to collaborate with AI agents effectively. Portugal, for instance, is noted for embracing GovTech partnerships, which can help bridge these gaps, but requires stable post-pilot funding to sustain progress.