Why planning needs to coordinate on AI

Vitus Cheung is a planner and economist at Savills. He studied at the UCL Bartlett School of Planning before recently graduating from the University of Westminster. This post is based on his dissertation research.

By now, everyone has used AI to some degree. Whether it is for summarising policy or just taking notes, AI has quickly become prevalent in our professional roles.

AI is also very much on the agenda for central government. From saving costs to speeding up service delivery, digital technology appears an attractive solution for addressing the tremendous problems planning is currently facing in the UK. At a national level, the government has started to trial AI for digitising planning documents[1] and augmenting planning decisions[2].  

The question is, do planning organisations need to get serious about responding together in a coordinated approach?

How AI is creating new ethical concerns for planning

The use of AI raises ethical concerns beyond that of other digital technologies. AI is not neutral. It is designed, developed and maintained by real individuals, with distinct characteristics and worldviews. Assumptions and compromises are always made in the design of AI. Mistakes, biases and a tendency to apply trends without attention to context are well documented. While human bias may be contained within a narrow area, the scaling of any single AI across many organisations makes the risk highly problematic.

AI is also notoriously opaque. The more complex, and by extension, capable an algorithm is, the more difficult it is to understand how it reaches specific outcomes. A considerable degree of technical knowledge is needed to even understand the terminology and concepts at play. Planning is intertwined with politics and law, subject to challenges and controversy. How can planning restructure a system of accountability that includes the widespread use of AI? These are just some of the many ethical aspects that need to be considered by planners. 

Why AI requires greater preparedness from planners

The founding principle of planners is to create better outcomes for people and places. Planners have a public purpose that requires attention to different needs of diverse communities, engagement on complex issues, and a duty to act as a mediator between impersonal forces and the very personal lives of individual humans.

If planners are to adopt a stewarding role when it comes to AI and communities, the separation between technical professions and planners cannot be maintained. Planners should understand how AI operates. They should help develop ways of working around AI’s limitations, embedding nuance, objectivity, transparency and accountability into emerging systems. Planning is already opaque and subjective in the public eye, but if planning becomes opaque and uncontrollable for planners themselves, planners will no longer be able to uphold a role as public stewards of our cities.

Coordination so far

For my recent Masters’ dissertation research I interviewed representatives from key planning organisations: MHCLG, the RTPI, local authorities and other stakeholders in digital planning. It is evident that there is currently minimal coordination between organisations. Dialogue remains confined to a small group of individuals, and focuses on the opportunity space and barriers to implementation. Planning as a whole is not aware of movements towards the integration of AI in planning, let alone serious consideration of ethical considerations or issues of preparedness. There is a disconnect between central government efforts to push technological solutions, and the practical limitations and understanding on the ground in local authorities.

Planning organisations are also currently unaware, unwilling or hesitant on the role their organisation could take on in directly addressing ethical considerations and preparedness. Guidelines are starting to appear for AI use, in appeals[3] and validation[4] for example, but there is clearly a long way to go for comprehensive, clear guidance at a planning-wide level. Likewise, there is no agreement on the level of skills and training necessary, with planning organisations each independently dealing with this (or not).

Recommendations

The following points represent a few of the primary areas of action identified.

1. Engagement with the solution space for the ethical design and use of AI. Areas of emerging research such as Ethical Frameworks and Explainable AI should be explored with a focus on adapting principles for the specific context of planning practice in the UK.
2. Build up a picture of readiness for the integration of AI in local authorities. Variance in the engagement and preparedness of local authorities towards AI and digital planning requires at minimum a national, standardised understanding of where capacity is not sufficient and what support is necessary from other planning organisations.
3. Consider the human dimension to changes brought about by AI and digital planning. For example, coordination should look at embedding motivated and innovative planners in local authorities as point contacts to manage change. Coordination should consider how information relevant to integrating new technologies can be collected, organised and made easily accessible and digestible, addressing the current lack of awareness. Perhaps most importantly, coordination should develop a robust shared vision for AI in planning that can be communicated in a succinct narrative and responds to the concerns of planners.

Planning at a crossroads

Perspectives towards the future of technology in planning range from scepticism to positivity. Planners often express faith in the public purpose of planning, while recognising the struggles the planning system currently faces. In this transition, planners have a window of opportunity to make AI work well for planners and planning. AI has significant potential to improve the ability of planning to create well-designed places and deliver lasting benefits for communities. Planners have an opportunity to be leaders in integrating AI in urban governance. But to do that, we will need to get serious about coordinating on AI.

[1] GDS Incubator for Artificial Intelligence, 2025. Extract.

[2] DSIT tender notice, 27 October 2025. MHCLG Augmented Planning Decisions.

[3] MHCLG, 2024. Use of artificial intelligence in casework evidence.

[4] North Norfolk District Council, 2025. Local Validation List 2025 to 2027.