AI‑Powered PlanAssist to Expand Across Twelve More English Councils After Pilot Success

PlanAssist, an AI‑driven planning tool built on DeepMind’s Gemini model, is being rolled out to twelve additional English local authorities after pilot success.

Developed by Google DeepMind in partnership with the UK government, Google Cloud and the consultancy Faculty, the system is designed to automate routine householder planning checks.

Householder applications, which include loft conversions, conservatories and small extensions, represent up to 70 % of a planning officer’s workload.

Despite their relative simplicity, these cases often create bottlenecks that delay housing projects and increase costs for families.

How PlanAssist Operates

The tool first validates submitted documents and architectural drawings to confirm that all required information is present.

It then cross‑checks the proposal against heritage constraints such as tree‑preservation orders and against both local and national planning policies.

PlanAssist also scans historic records for prior objections or issues raised by residents or other stakeholders.

Finally, the system drafts a fully cited assessment report that planning officers can edit before issuing a decision.

Pilot Outcomes and Scaling Plan

Early pilots in Barnet, Camden and Dorset councils demonstrated a 58 % reduction in the average time officers spent on householder assessments.

More than 70 % of the automatically generated reports were approved by officers without any edits, indicating high accuracy.

Since June, the live‑application beta has continued to confirm these efficiency gains across the three pilot councils.

The next phase will introduce PlanAssist to twelve further councils in three cohorts, bringing the total participating authorities to fifteen.

These new councils span both metropolitan and rural districts, allowing DeepMind to test the tool’s performance in diverse planning environments.

Subject to successful testing, the UK government intends to make PlanAssist available to all English councils by 2027.

Even with AI assistance, every assessment will still be reviewed and approved by a qualified planning officer before any decision is finalized.

Implications for Housing Delivery and Public Services

Accelerating routine checks helps address the UK government’s target of delivering 1.5 million new homes by 2029.

Faster approvals can reduce waiting periods for families seeking modest home improvements, improving quality of life.

By freeing officers from repetitive tasks, the system enables them to focus expertise on more complex or contentious applications.

The collaboration illustrates how public‑sector AI deployments can complement human judgment rather than replace it.

DeepMind’s partnership with the government also signals growing confidence in using advanced models like Gemini for civic operations.

Continued monitoring will be essential to ensure that AI recommendations remain transparent, unbiased and aligned with local planning objectives.

Overall, PlanAssist showcases a pragmatic approach to AI adoption in public administration, balancing efficiency gains with human oversight.

Why This Matters: The rollout means planning officers across England will handle applications faster, potentially easing housing shortages while preserving expert decision‑making.