AI in UK Planning: What Farmers Need to Know in 2026
Planning officers are using AI to process your application. This is not a future development. It is happening now. Google's Gemini model is already in use across English councils, and a second tool that makes grant-or-refuse recommendations on applications is currently being piloted. If you submit a farm planning case in 2026, there is a real chance AI will touch it before a human does.
Here is what that means for you.
What the Government Has Actually Deployed
There are two distinct tools in play. It is worth understanding both.
Extract: turning paper records digital
The UK Government's AI Incubator team developed a system called Extract, built on Google DeepMind's Gemini model. Its purpose is narrow but significant: it takes old, handwritten planning documents and converts them into digital form in seconds — jobs that would otherwise have taken hours.
Extract uses Gemini's advanced visual reasoning and multimodal capabilities to help councils turn old planning documents — including blurry maps and handwritten annotations — into structured, usable data.
The system was piloted with four councils — Hillingdon, Westminster, Nuneaton and Bedworth, and Exeter — and is being rolled out to all councils in England through 2026. The aim is to have every local planning authority equipped to digitise all their planning documents by the end of the year.
Why does this matter for farm applications? Because every year, councils must process around 350,000 planning applications, many of which involve legacy paperwork that is time-consuming to interpret. Extract clears that backlog. It also means that historic planning decisions on your farm — including old prior approvals, agricultural condition notices, and enforcement records — are becoming searchable. Officers will be able to find them faster than ever.
If there is anything in your farm's planning history that a case officer should know about, assume they will find it.
The Augmented Planning Decision Tool: AI recommending grant or refuse
This is the development with more direct implications for applicants.
A Google-built AI system, the Augmented Planning Decision Tool, is being piloted by English councils to generate grant-or-refuse recommendations on building applications, with officers retaining final sign-off.
Google won an £8.3 million contract to build a tool that produces reasoned, verifiable recommendations and writes draft reports. It will initially focus on householder applications, with the goal of reducing average decision times from eight weeks to four.
The platform assists planning officers by carrying out policy research, generating citations from planning regulations, and checking whether proposals comply with existing rules. The officer reviews the AI's output and makes the final decision. The AI does the groundwork first.
What This Means for Agricultural Permitted Development
The Augmented Decision Tool is currently focused on householder applications — extensions, loft conversions, and the like. It is not yet processing Class Q, Class R, or Part 6 applications. However, there are two reasons not to treat this as reassuring.
First, the government's stated direction of travel is clear. The AI tool is designed to cut processing times in half. If it works for householder cases, the political pressure to extend it to other application types will be substantial. Agricultural permitted development — already a high-volume, policy-intensive area — is an obvious candidate.
Second, Extract is already live and expanding. Once your farm's full planning history is digitised and searchable, an AI system checking policy compliance will be working from a much richer dataset than a case officer reviewing a paper file. Historic decisions, condition registrations, and prior approval records will all be in play.
In our work with landowners across England, we already see officers cross-referencing planning histories in detail. AI makes that faster and more thorough, not less.
Why Application Quality Has Never Mattered More
There is a common assumption that AI will make planning easier — that faster processing means less scrutiny. The opposite is likely true.
An AI system that checks policy compliance will do so consistently. It will not miss a paragraph reference. It will not overlook a condition. A poorly drafted agricultural prior approval application — one that misstates the legislative basis, omits a required assessment, or makes an argument that does not track the relevant Class criteria — will be flagged efficiently.
The margin for error in application quality is narrowing, not widening.
Planning decisions require professional expertise to consider subjective, creative, and complex considerations. AI tools have a helpful part to play, but they are no substitute for professional advice, common sense, and informed judgement. That is where the value of a specialist remains.
AI will handle the administrative and policy-checking layer. The professional argument — how a building is characterised, how functionality is demonstrated, how structural integrity is evidenced — still requires a specialist.
What Farmers Should Do Now
If you are considering a Class Q conversion, a Class R change of use, or any other agricultural prior approval, this shift in how applications are processed has practical implications.
Your application needs to be watertight on the policy framework from the outset. The GPDO 2015, as amended by SI 2024/579, sets out specific criteria for each Class. An AI checking compliance will apply those criteria literally. Ambiguity in how your submission addresses them is a risk you can manage before submission — not after.
Your farm's planning history matters more than it used to. Before submitting, it is worth understanding what is already on the register. Conditions, prior approvals, and enforcement records can all affect the viability of a new application. We carry out portfolio-wide planning audits for this reason.
And the officer who reviews the AI's recommendation still has professional discretion. The quality of the supporting case — the planning statement, the structural report, the design and access content — will influence how that discretion is exercised.
Frequently Asked Questions
Will AI make planning applications faster for farmers?
Potentially, in the long run. The current focus is on householder applications. Agricultural permitted development involves more complex policy assessment, and there is no confirmed timeline for AI tools to be extended to Class Q, Class R, or Part 6 cases. The more immediate effect is that document processing and policy-checking will become quicker and more thorough across the board.
Can AI refuse my planning application?
No. Officers retain final sign-off. The AI produces a recommendation; the case officer makes the decision. The professional judgement of the officer — and the quality of the case you put before them — still determines the outcome.
Will my farm's planning history be visible to AI tools?
Yes, over time. The government's aim is to have every local planning authority equipped to digitise all planning documents by the end of 2026. That includes historic prior approvals, agricultural condition registrations, and enforcement records. Once digitised, that data feeds into the AI tools being deployed at officer level.
Does this affect Class Q applications submitted now?
Indirectly. Extract is already operational in pilot councils and expanding. The Augmented Decision Tool is focused on householder applications for now. But the direction of travel is clear, and applications submitted today will sit in systems increasingly shaped by these tools. Quality of submission matters regardless of which tool reviews it.
What is Google Gemini's role in UK planning?
Extract is built on Google DeepMind's Gemini model, running via Google Cloud's Vertex AI platform. It reads and analyses complex documents, pulling critical information from text, images, and handwritten annotations. The Augmented Planning Decision Tool is also a Google Cloud product, built under an £8.3 million government contract.