Can AI Finally Fix the UK’s Slow and Outdated Planning System?

Is Britain’s painful, paperwork-choked planning process about to get revolutionized by artificial intelligence?
For decades, UK councils have struggled with mountains of paper maps, handwritten notes, and slow bureaucracy—delaying crucial housing and infrastructure projects. But a bold new tool built with AI, called Extract, is promising to flip the script, making approvals faster, fairer, and digital-first. Could this be the moment where tech truly transforms how cities are built? Let’s dive in.

🗂️ The Age-Old Planning Headache: Outdated, Overwhelming, and Onerous

For companies, councils, and citizens, the planning system has long meant headaches:

• Hundreds of thousands of planning applications: Nearly 350,000 planning documents are processed in England every year, often handwritten or low-quality scans.
• Manual bottlenecks: Reviewing and digitizing just one document can take a professional up to 2 hours—multiplied by the hundreds of thousands stored in filing cabinets across the country.
• Glacial timelines: These stacks of paperwork slow down everything, from home improvement permits to billion-pound infrastructure projects.
• Mountains of legacy data: Blurry maps, scribbled notes, and old blueprints lock away essential information, preventing modern digital tools from making sense of them.

The underlying problem? The UK’s planning system is still heavily reliant on paper and manual review—a reality that drives up costs, saps productivity, delays economic growth, and locks up information that could unlock progress.

🚀 Enter Extract: The AI-Powered Game Changer

So what’s being done? Enter Extract, the UK's new AI-driven platform built using Google’s Gemini model via Vertex AI. Here’s why the tech world—and government leaders—are excited:

• Unmatched multimodal muscle: Gemini doesn’t just read typed words—it reasons about images, understands handwriting, and can even identify features on old, fuzzy maps.
• Lightning-fast results: What once took 2 hours now takes just 40 seconds—thanks to Extract’s ability to digitize and extract critical information from legacy planning documents.
• Real-world pilot programs: Four pioneering councils—Hillingdon, Westminster City, Nuneaton and Bedworth, and Exeter—are already putting Extract to the test.
• Scaling up: By Spring 2026, every UK council will have access to Extract, democratizing digital planning tools nationwide.

Here’s how it works, broken down:

• Polygon Extraction: Gemini spots critical map features—like property boundaries or restricted zones—using visual reasoning and advanced tools (OpenCV, Ordnance Survey, Segment Anything).
• Georeferencing: Addresses, street intersections, and landmarks are identified and matched with modern digital maps, transforming old paper borders into precise GPS coordinates using state-of-the-art models like LoFTR.

The promise? Faster housing approvals, reduced council workload, greater transparency, and ultimately, a nimbler, 21st-century planning system fit for a modern Britain.

✅ The Benefits: How Extract and AI Can Unlock Progress

• ✅ Supercharged efficiency: Transforming 2-hour data entry into a 40-second AI process frees up countless work hours for planning professionals.
• ✅ Faster decision-making: Councils can approve applications and update restrictions much faster, supporting home building and infrastructure upgrades.
• ✅ Democratized information: Digitized planning data means the public and professionals can access clear, reliable information online—cutting out confusion.
• ✅ Strategic focus: With the grunt work automated, planners can focus on shaping smarter, long-term strategies for their communities.
• ✅ Scale for impact: If fully rolled out, Extract could help deliver on the government’s pledge to build 1.5 million new homes and cut red tape for infrastructure projects nationwide.

🚧 Challenges and Questions: Will AI Deliver or Disappoint?

• 🚧 Legacy data headaches: The sheer volume and variability of old maps, scrawled notes, and document types present tough conversion challenges.
• ⚠️ Accuracy and oversight: Critical planning decisions require extreme precision; automated extraction must be meticulously checked to avoid costly mistakes.
• 🚧 Digital adoption at scale: Councils are at different stages of digital readiness, meaning Extract’s rollout could face hurdles—or even resistance—in less tech-savvy regions.
• ⚠️ Training and trust: Planners and the public will need reassurance that AI is a tool for empowerment—not error or exclusion.

Expert voices:

• UK Prime Minister Keir Starmer calls Extract “a bold step forward in our mission to build 1.5 million more homes and deliver a planning system that’s fit for the 21st century.”
• Google DeepMind CEO Demis Hassabis says it’s “exciting to see the UK government choose Gemini to help speed up the planning process and support planners and people across the country.”
• Local authority leaders are already reporting improved speed and geospatial data accuracy during pilots.

🔭 Final Thoughts: Revolutionary, But Only If Implementation Is Smart

• ✅ If Extract delivers on accuracy and councils embrace digital, expect faster homes, less red tape, and empowered communities.
• 📉 If legacy systems or trust issues slow adoption, the UK risks missing this golden opportunity for public service innovation.
• 🚀 With continued government and private sector support—including Google Cloud’s hosting, AI engineering, and ongoing pilots—the sky’s the limit for AI-powered transformation in planning… and beyond.

Is this the dawn of a bold, digital-first era for UK planning? Or will the real world of legacy systems and local resistance prove too tricky for even AI to solve? What do you think—can Extract make British planning better, faster, and fairer? Let us know in the comments!

Let us know on X (Former Twitter)

Sources: Owen Larter. UK government harnesses Gemini to support faster planning decisions, June 2025. https://blog.google/around-the-globe/google-europe/united-kingdom/uk-government-harnesses-gemini-to-support-faster-planning-decisions/