Leeds reports more planning applications processed with Xylo

AI system

System

Xylo Core AI planning workspace

Model

The council listed GPT-5, Gemini 2.5 Pro, and Claude Sonnet 4.

Editor’s note. The council’s disclosure dated 27 November 2025 lists these models. It does not identify which model produced each reported result.

Release

The council’s disclosure of 27 November 2025 lists gpt-5-2025-08-07, gemini-2.5-pro, and claude-sonnet-4@20250514. It does not connect individual results to one of those releases. [3]

Application developer

Xylo

Role

The models retrieved planning information and generated suggestions and draft text for officers to review.

Configuration

The council’s November 2025 disclosure listed gpt-5-2025-08-07, gemini-2.5-pro, and claude-sonnet-4@20250514.

Inputs and outputs

Not assessed

Autonomy

Not assessed

Workflow

Leeds and Xylo co-designed and tested the workspace over more than six months. The reported pilot focused on time, work quality, and officer wellbeing.

Human oversight

Not assessed

Relation to AI

The council describes retrieval of relevant planning material and API calls to language models that generate suggestions and draft text for officer review. Its November 2025 disclosure identifies GPT-5, Gemini 2.5 Pro, and Claude Sonnet 4. The models do not determine planning merits. Reported changes in work and wellbeing come from project-side interviews and usage data. A supplier-reported control group lacks sufficient published methods to establish an independent causal comparison.

AI contribution

Benefits

Outcome identifiers: Not assessed
Benefit: Leeds reports over 40% more applications processed per officer, less research time, and improved officer wellbeing during its Xylo pilot.
Category: Improved Organizational Efficiency
Status: Reported benefit
Beneficiary groups: Organizations. People.
Beneficiaries: Not assessed
Reach: Not assessed
Distribution and access: Not assessed
Duration: Not assessed
Expected?: Yes
Costs and harms: Not assessed

Measurement

Studies: Not assessed
Study design: Not assessed
Measurement: Not assessed
Setting and sample: Not assessed
Measures: Not assessed
Comparison: Not assessed
Results: Not assessed
Additional value from AI: Not assessed
Limitations: Not assessed
Denominators: Not assessed
Materials for replication: Not assessed

Money

Kind of value: Not assessed
Amount: Not assessed
Currency and prices: Not assessed
Period and unit: Not assessed
Realization: Not assessed
Recipients and payers: Not assessed
Gross or net: Not assessed
Included costs: Not assessed
Comparison: Not assessed
Method: Not assessed
Return and payback: Not assessed

People and places

Participants: Developer: Xylo. Operator: Leeds City Council.
Jurisdictions: Benefit: United Kingdom.
Ownership and control: Not assessed
Dependencies: Not assessed
Cross-border connections: Not assessed
Geopolitical implications: Not assessed

Follow-up

Conditions for replication: Local planning policies and application material are available and kept current for retrieval. Planning officers review generated suggestions against the original material and retain responsibility for decisions. Integration limits, officer training, and the time required for review affect the achievable gains.
Replication attempts: Not assessed
Later findings: Not assessed
Availability: Not assessed
Institutional decision: Not assessed
Grounds: Not assessed
Action and result: Not assessed
Success criterion: Not assessed

Evidence

Source numbers: AIBD-0553-source-0001–0004
Citations: Numbered references below. [1] [2] [3] [4]
Language: Not assessed
Source relationships: Operator, Developer [1] [2] [3] [4]
Original evidence: The council’s account has no independent evaluation or public dataset of outcomes. The technical overview, complete internal assessments, and pilot calculations were inaccessible. [1] [2] [3] [4]
Access and versions: Not assessed
Supported findings: Not assessed
Evidence links: Not assessed
Evidence strength: Not assessed
Missing evidence: Not assessed

Grades

Assessed benefits: Organizations. People.
Depth: Organizations: 2. People: 2.
Reasons: Organizations: Reported application throughput rises by over 40%, supporting a measurable service-capability gain. People: All surveyed officers reportedly experienced improved stress and energy levels, alongside less research work. These are reported wellbeing/work-capability improvements without a demonstrated lasting change.
Observed or intended: Reported benefit
Rubric: Not assessed
Assessor and date: Not assessed
Alternative assessments: Not assessed
Fill: The deployer's account

Entry details

Kind: Not assessed
Related entries: Not assessed
Field: Planning administration.
Purpose: Help planning officers find and process relevant information while retaining professional judgment.
Dates: Source published: 12 March 2026. Updated: 24 October 2026.
Stage: Not assessed
Version: SHA-256: 547cdd694349…

References

Other identifiers: LGA AI case-study bank. LGA-013.
Category mappings: Not assessed
Brief: Not assessed
Revisions: 2 revisions. Latest: 24 October 2026.

Sources

Report Deployer

Leeds City Council and Xylo, published by the Local Government Association. Leeds City Council and Xylo: transforming planning with AI. 2026-05-08. Sections “The collaborative process,” “Successes and results,” and “Risk identification and mitigation.” Original project account with pilot participation and outcome figures; no underlying evaluation dataset or independent assessment.

deployer
Report Developer

Habs Isabelle Kim, Xylo cofounder. Leeds planning pilot: 40% more applications than the control group. 2026-03-12. Original supplier account of results presented alongside Leeds planning officer Benedict Barnett. Page metadata gives March 12, 2026, 11:05:31 UTC. Adds the control-group claim and earlier chronology, without disclosing methods.

makers
Report Deployer

Leeds City Council and Xylo. Leeds City Council: Xylo Core. 2025-11-27. Algorithmic Transparency Recording Standard, §§2.1, 3.2–3.3, 4.1–4.2, and 5.1–5.2. Original council/supplier disclosure of system design, named model versions, human review, and risks. Publication on GOV.UK is not independent regulatory confirmation of the outcomes.

deployer
Report Developer

Xylo. Leeds Patch pilot: six officers and 33 pre-applications. 2026-03-11. Original supplier report describing a pre-application pilot and its sample. Retained to distinguish that sample from the Core householder/minor throughput result. The post is marked edited and provides no public version history.

makers