AI use cases in local government: concrete examples

According to the Observatoire Data Publica barometer (November 2024, relayed by La Banque Postale), around 36% of French local authorities have already put in place or trialled an artificial intelligence system, and ~15% say they intend to take the step within the next 12 months. The documented use cases show that results are tangible where the tool answers a specific business process, not a general technological goal.

Overview: use cases by function

Citizen relations and administrative procedures

Citizen relations is the most visible area of application — and often the first to be trialled. It mainly involves automating answers to frequent questions, directing citizens to the right services, or handling incoming requests more quickly.

The town of Plaisir (Yvelines) deployed an AI assistant to manage incoming requests from citizens, with documented results on reducing unanswered messages (source: Smart by Design).

The town of Issy-les-Moulineaux launched "IssyGPT", an AI-powered assistant that handled nearly 20,000 citizen queries in under a year (source: Smart City Mag, 2024).

The bespoke approach: a generic chatbot plugged into a consumer language model poses a real risk for a local authority: inaccurate answers, citizen data sent to a third-party service, and no control over what the tool says on its behalf. The alternative is to design a tool trained on the authority's own data, hosted under its control, and able to escalate to a human agent. That's what Groupe Milestone builds: business applications that integrate AI as a functional layer, not as an external product.

Human resources and internal management

The HR departments of local authorities handle a large volume of documentation: meeting minutes, correspondence, internal notes, schedules. AI can reduce the administrative load on these tasks to free up time for higher-value work.

The town of Suresnes put in place an automatic meeting transcription tool, producing structured minutes without manual re-entry (source: Envies de ville, Safebrain).

The bespoke approach: transcription is a basic building block — but what gives a business HR tool its value is its ability to fit into the existing workflow: automatically feeding the register of decisions, extracting the decisions taken and assigning them to the right services, flagging follow-up deadlines. A bespoke tool can string these steps together into a single journey.

Finance, purchasing and public procurement

Document analysis is a use case particularly well suited to AI in finance and procurement services: reading and comparing lengthy bids, detecting anomalies in accounting data, keeping legal watch on regulatory texts.

The town of Cannes documented eleven AI use cases, including analysing public procurement bids, translating documents, transcription and legal monitoring (source: Hypra, local government AI benchmark 2026; Safebrain).

The bespoke approach: analysing procurement bids is a sensitive task. A bespoke tool built for this use case incorporates the authority's business rules, the analysis criteria specific to each procedure, and produces verifiable summaries — with a human agent remaining the decision-maker.

Technical services, roads and public assets

Technical departments manage physical assets whose maintenance is costly and whose failures are unpredictable. AI, combined with sensors or inspection data, enables a more anticipatory approach to management.

• Predictive asset management: analysing fault histories to anticipate preventive interventions.
• Damage detection: automatic analysis of road or building inspection photos (potholes, cracks) to prioritise interventions.
• Route optimisation: automatic calculation of collection or maintenance routes based on field constraints.

An IoT touch: for some projects, the data comes from the field — temperature sensors in buildings, fill-level probes on containers, presence detectors. Groupe Milestone supports projects that integrate connected objects (IoT) into a data collection and analysis setup linked to a bespoke business tool. On the sovereignty of this data, see the data sovereignty section of the pillar article.

Environment, energy and risk prevention

Local authorities face growing obligations on environmental monitoring and risk prevention. AI can help analyse continuous data streams, detect anomalies and trigger alerts.

• Energy monitoring of public buildings: aggregating consumption data, automatically detecting overshoots, recommending optimisations.
• Natural risk prevention: cross-referencing meteorological, hydrological and geographical data to anticipate risks and trigger alerts.
• Water quality management: analysing continuous sensor data to detect anomalies and ensure regulatory compliance.

From example to your tool: the difference bespoke makes

The examples above share one thing in common: they worked because they answered a specific business process, with identified data, defined users and measurable results.

What sets a successful project apart from a disappointing one isn't the AI technology — it's the quality of the design work upstream. A generic assistant deployed quickly can produce spectacular results in a demo, and disappointing ones in real use. A tool designed for your exact process, with your data, your business vocabulary and your validation rules, produces reliable, usable results from the moment it goes live.

That's the difference between AI as a product and AI as the augmentation of a bespoke business tool.

Discover our services: Automation & AI · Business applications.

To take action within the framework of public procurement: Having a bespoke AI business application developed for your authority.

Examples from third-party local authorities are cited with their sources and do not constitute Groupe Milestone references.