The Borough of Broxbourne Council, serving approximately 100,000 residents in Hertfordshire north of London, has deployed Jadu's AI-powered search technology Agent-Ex: Search. The tool uses natural language processing to help citizens find information on the council's citizen portal more efficiently than traditional keyword-based search.
Agent-Ex: Search interprets user queries in plain English and returns contextually relevant results from the council's content library, including service descriptions, policy documents, and procedural guidance. The system is designed to reduce time spent navigating complex council websites and to lower the volume of routine enquiries handled by customer service staff.
AI adoption in UK local government accelerates
The Broxbourne deployment is part of a broader trend in UK local government. More than a dozen councils have begun testing or implementing AI-driven tools for citizen-facing services since 2024, driven in part by budget constraints and the UK Local Digital Fund which co-finances digital transformation projects. The Department for Levelling Up, Housing and Communities has prioritised automation of routine administrative processes to free up staff capacity for complex case work.
Jadu, a UK-based provider of digital experience platforms for the public sector, has been positioning Agent-Ex as a low-risk entry point for councils exploring administrative automation. Unlike generative AI chatbots that produce new text, Agent-Ex: Search retrieves and ranks existing, pre-approved content, reducing the risk of inaccurate or misleading responses. The system does not make decisions or process transactions; it functions as an intelligent index.
Implications for administrative digitalisation
The technology addresses a persistent challenge in e-government: information accessibility. A 2025 study of KI deployment in the public sector found that citizens abandon more than 40 per cent of online service requests due to difficulty finding the correct form or guidance. AI-driven search can reduce that drop-off by interpreting colloquial or vague queries and surfacing the relevant service page.
However, the effectiveness of such tools depends on the quality and structure of underlying content. Councils with outdated or fragmented information architectures may see limited benefit. Jadu has recommended that councils conduct content audits and tag documents with structured metadata before deploying Agent-Ex, a step that requires dedicated staff time and editorial discipline.
The Agent-Ex rollout at Broxbourne will be monitored for six months, with metrics including search success rate, time on site, and referral patterns to customer service channels. If the pilot proves effective, Jadu expects further adoption among borough and district councils facing similar resource pressures. The UK Government Digital Service strategy for 2025–2027 explicitly encourages local authorities to share AI tooling and evaluation frameworks, signalling policy support for incremental automation.
For procurement officers and digital leads in UK councils, the Broxbourne case offers a concrete reference point: a defined scope, manageable integration effort, and measurable outcomes. It also highlights the importance of aligning AI tools with existing content governance processes rather than treating them as standalone technical fixes.
