Councils face budget pressure, democratic accountability, and a workforce starting from very different places. AI adoption here requires a fundamentally different approach from commercial settings. Mike has led this work from the inside — as a Cabinet Member with direct portfolio responsibility for AI strategy and digital services.
The challenge isn't reluctance. It's that the governance environment is genuinely more complex — and that complexity is not optional.
Budget pressure is driving genuine interest in AI across local government. Councils are looking at tools to do more with less. But unlike commercial organisations, every decision must be defensible. An AI initiative that generates cost savings is not enough on its own — the same initiative must withstand scrutiny from opposition councillors, overview and scrutiny committees, audit teams, and the public.
Democratic accountability means transparency. AI systems used by councils affect residents directly. A council cannot simply adopt a tool and operate it quietly. Equalities impact assessments are not optional extras — they're fundamental to lawful operation. GDPR compliance takes on a different character in a public context. Data breaches aren't just commercially damaging; they're scrutinised by commissioners and the press.
Procurement frameworks add real constraints. Many councils use frameworks like G-Cloud or Crown Commercial agreements, which limit tool choices and require audit trails that commercial SaaS providers may not naturally support. Union considerations matter in ways they often don't in the private sector.
And the workforce is genuinely diverse in its digital confidence. An AI pilot might reach officers who are deeply comfortable with digital tools, but it will also need to land credibly with elected members who may be entirely new to any of this. When a resident complains to their councillor about an AI-assisted decision, the council needs to be able to explain it clearly. That requires a different kind of governance than pure efficiency.
Mike doesn't advise councils from the outside. He's done this work from the inside — as the elected Cabinet Member with portfolio responsibility for Digital Services and Climate at Cotswold District Council.
This isn't consultant-grade familiarity. It's direct accountability for AI adoption decisions within a democratic local authority. The work included:
Responsible AI Policy: Co-authored the council's formal Responsible AI Policy — setting out the governance framework, ethical principles, transparency requirements and accountability structures for all AI use across the authority. This was ratified at Cabinet level and provides the foundation for defensible AI deployment.
AI Strategy & Adoption Governance: Co-authored the council's AI Strategy, establishing the adoption governance framework including an AI Steering Group, approval processes for new use cases, and clear escalation routes. This gave officers a structured path from experimentation to sanctioned operational use.
Microsoft Copilot Rollout: Oversaw the deployment of Microsoft Copilot to approximately 250 officers and elected members — one of the earlier council-wide AI tool deployments in English local government.
Planning — Development Management: AI assistance for planning officers processing applications, including automated summarisation of submissions, policy compliance checking, and draft officer reports. Reducing the time from application to decision in a service under significant volume pressure.
Planning — Local Plan Development: AI-assisted analysis of consultation responses, policy research synthesis, and evidence base preparation for the emerging local plan. Handling the volume of public input that would otherwise require weeks of manual review.
Planning — Retrofit & Residents' Homes: AI tools supporting the assessment of residential properties for energy retrofit, helping officers identify suitable properties and match them to available grant funding and intervention programmes.
Democratic Services: AI assistance with minute-taking, agenda preparation, and document summarisation for committee meetings — reducing administrative burden on democratic services officers while maintaining the accuracy and formality required for public record.
Communications & Customer Services: AI-assisted drafting of public communications, press releases, and resident correspondence. Self-service AI tools to help residents find information and resolve queries without needing to contact the council directly — reducing call volumes while improving response times and consistency.
England's Local Government Reorganisation programme is creating new unitary authorities from 2026 onwards. Most councils are focused on the mechanics of merging. The real opportunity is designing the new authority with AI built into its operating model from day one.
LGR replaces England's two-tier council structure — county and district councils — with single-tier unitary authorities. Over twenty areas are affected, with shadow authorities forming in 2026–27, elections in May 2027, and new councils going live from April 2028. Every affected council faces the same core challenge: merging multiple organisations with different systems, processes, cultures, and service standards into a single authority that must deliver from day one.
Most reorganisation programmes focus on harmonisation — aligning pay, terms, IT systems, and processes. That's necessary, but it's the minimum. Councils that treat LGR purely as a consolidation exercise will emerge with a larger version of the same operating model. The services will be merged, but not redesigned.
The alternative is to treat LGR as what it actually is: a rare opportunity to design a new council from scratch. And that means designing it with AI as part of the workforce from the outset, rather than trying to retrofit AI into inherited ways of working.
Councils going through reorganisation are already rebuilding their processes. The question is whether they rebuild them the old way at a larger scale, or redesign them for a fundamentally different operating model. The councils that get this right will:
Design new citizen-facing services with AI embedded — not bolting chatbots onto legacy systems, but designing customer journeys where AI handles triage, self-service and routine queries from the start, with clear escalation to human officers for complex cases.
Use AI to accelerate the transition itself — mapping duplicate processes across merging authorities, identifying harmonisation opportunities, and bridging incompatible systems during the migration period.
Build the new operating model around hybrid workflows — defining roles, governance and oversight structures that treat AI as part of the workforce rather than an afterthought. Getting this right at inception is dramatically easier than retrofitting it later.
Establish AI governance from day one — embedding responsible AI principles, approval processes and monitoring into the new authority's constitution and operating procedures, rather than layering them on after the fact.
The AI opportunity in local government is significant — but it's concentrated in specific areas. Understanding which those are matters before investing in tools or programmes.
Planning is one of the highest-volume, most document-intensive services in any council. AI can summarise applications, check policy compliance, draft officer reports, and accelerate the path from submission to decision. At Cotswold, pilots in development management showed how AI assistance could meaningfully reduce processing time without compromising quality or accountability.
Residents contact councils with questions that often have straightforward answers buried across multiple web pages and documents. Well-designed AI self-service can resolve a significant share of these queries instantly, reducing call volumes while improving response consistency. The governance challenge is ensuring AI responses are accurate and that complex or sensitive cases always reach a human officer.
Correspondence handling, document summarisation, report drafting, casework research. These are high-volume, repetitive, and low-governance-risk. A council officer spending two hours on routine correspondence can often reclaim that time with good AI assistance. The ROI is clearer and the governance questions simpler — you're accelerating existing processes, not creating new ones.
Committee agendas, minutes, forward plans, member briefings — democratic services teams manage a constant flow of formal documentation under tight deadlines. AI can draft minutes from recordings, summarise lengthy reports for member briefings, and help prepare agenda packs. This reduces administrative burden while maintaining the formality and accuracy the public record demands.
Press releases, consultation communications, social media content, resident correspondence — councils communicate constantly and consistency matters. AI can help draft, review and adapt communications across channels while maintaining the council's tone. At Cotswold, AI-assisted communications reduced drafting time and improved consistency across the authority.
Senior officers and elected members need to absorb large amounts of information quickly. AI can synthesise consultation responses, prepare policy briefings, and summarise evidence bases. The governance questions here are about audit trail and attribution rather than direct resident impact — but getting the analysis right matters enormously for the decisions that follow.
Every council has different starting points, different political contexts, and different service pressures. The engagement always starts with understanding yours.
Whether you're a district council exploring your first AI use case, a county authority preparing for LGR, or a new unitary designing its operating model — the approach is structured around your governance reality, not a generic playbook.
A focused 2–4 week assessment that maps where AI can genuinely reduce cost or improve service delivery, identifies which governance requirements must be resolved first, and produces a realistic pilot recommendation. This isn't a consultant's wish list — it's structured discovery of genuine opportunity against genuine constraint. You emerge with a clear workflow opportunity map, a governance risk assessment, and a prioritised action plan.
For councils approaching LGR, the Radar can be specifically focused on identifying where AI should be embedded into the design of the new authority's operating model.
A 6–12 week pilot in a real operational context — not a proof of concept or a sandbox. A genuine working pilot where governance is built in from day one. You're testing how AI actually behaves in your workflows, training your team, and building the audit trail and risk controls that let you scale with confidence. For councils, this typically means working with a single service area and embedding an officer-led pilot that generates both usage data and governance evidence.
A 3–6 month engagement to move from pilot success into operational sustainability. This is where governance really matters. You're embedding AI into your operating model — defining roles, establishing monitoring, building training pipelines, and creating structures that let councillors and officers understand and defend the decisions being made. For councils with multiple service areas, this often means establishing a shared approach and clear escalation routes across the authority.
For new unitary authorities, this is the critical engagement — designing the AI-ready operating model for the new council before legacy habits become the default.
A retained engagement — typically 1–3 days per month — providing senior AI oversight. Especially valuable for councils where AI decisions span multiple service areas or where you need board-level credibility without a permanent hire. Mike can serve as your fractional head of AI, provide governance advice, support procurement decisions, and help ensure council AI deployment stays on track and remains politically defensible.
Governance-grounded, practically focused AI adoption — from someone who's done this work from the inside.