Development of AI for Public Engagement and Consultations
At the heart of our planning system is a commitment to consult with local residents, businesses, community organisations, and other stakeholders. This collaborative approach makes sure that the development plans for Greater Cambridge reflect the needs and aspirations of our community.
From the emerging joint Local Plan to site-specific Area Action Plans, Supplementary Planning Documents, and Neighbourhood Plans, public consultation is a mandatory and integral part of our planning process.
Developing AI to improve efficiency in our consultation process
In our most recent Local Plan consultation, the Greater Cambridge Shared Planning Service (GCSP) received over 9,500 responses. These submissions ranged from professionally prepared documents by commercial organisations to handwritten notes from residents. Each response is valuable and contributes to shaping our community's future.
To enhance the efficiency and effectiveness of processing these submissions, we secured funding from the central government’s PropTech Innovation Fund. This funding has enabled us to develop a cutting-edge AI tool in partnership with the University of Liverpool.
This tool, known as a Large Language Model (LLM), is specifically designed to process and summarise the thousands of comments received during planning consultations.
AI Tool development and testing
The development of the AI tool involved training it on over 100,000 submissions received by GCSP between 2012 and 2024, covering 164 draft planning documents and frequently used planning terminology. This bespoke training ensures the tool is finely tuned to both planning policy and the unique context of Greater Cambridge.
In January 2025, we tested the AI tool on 3 live public consultations. The tool generated summaries of each submission, which were then compared to those manually created by planning officers. The results showed no significant differences in quality, with the AI tool providing accurate and comprehensive summaries.
Efficiency gains
The AI tool demonstrated remarkable efficiency, summarising all submissions in just 16 minutes compared to the 18.5 hours required by human planners. This significant time saving exceeded our initial expectations and highlights the potential for AI to streamline the planning process. You can see a sample of the comparison between what the planners produced and what the AI tool produced in this evaluation document [PDF, 0.1MB].
Community feedback and ethical considerations
We value the input of our community and planning stakeholders. In early 2025, we conducted focus groups to gather feedback on the use of AI in the planning consultation process. The discussions covered ethical use, environmental impact, development process, and the reliability of AI-generated summaries. This feedback has been instrumental in refining the tool and our processes.
Read the focus groups outputs report [PDF, 0.4MB]
Read the frequently asked questions document [PDF, 0.1MB]
Ongoing commitment to quality and transparency
It is important to note that planners will continue to read each individual submission. The AI tool serves as an administrative support, not a decision-maker. Planners will retain the responsibility for making informed decisions based on the comprehensive review of all submissions.
Together, we are embracing innovation to build a better future for Greater Cambridge.
Future applications
While the AI tool is currently tailored to the systems used by GCSP, it has the potential for broader application with minor adjustments.
For more information, please contact:
Professor Alex Lord (University of Liverpool) - alexlord@liverpool.ac.uk
Terry de Sousa (GCSP) - terry.desousa@greatercambridgeplanning.org