Can AI help the government overcome the planning backlog?

Artificial Intelligence is rapidly transforming every facet of contemporary life. In the UK, the government sees AI as a game-changing tool to accelerate the planning process, boost housing development, and reshape urban environments. But how close are we to this vision? 

What are the challenges to overcome? To explore these questions, experts from UCL's Bartlett School of Planning—Professor Lauren Andres and Professor Mike Raco—have turned to cities like London and Beijing to study how AI is influencing urban planning systems, real estate development, and citizen engagement.

The Potential for AI to Transform Urban Planning

The UK faces an urgent need not only to build more homes but also to enhance the quality of housing developments. The government has set an ambitious target of delivering 1.5 million new homes during the current parliamentary term (2021–2026). However, a report by Savills (May 2025) reveals that only around 225,000 homes received full planning consent in the year to March 2025—significantly below the estimated 367,000 homes per year required to meet long-term housing demand.

As others have made celar, poor-quality housing—particularly in disadvantaged areas—remains the norm, and this must change.

In response, the UK government views the digital transformation of planning systems—such as the adoption of map-based tools and streamlined online applications—as a key strategy to accelerate development and promote ‘beautiful, sustainable, and locally-led communities’. For planners, especially those working within under-resourced local authorities facing severe financial constraints, these innovations are seen as both timely and transformative. Digital tools have the potential to reduce bureaucratic burdens, expedite decision-making, and improve access to critical information.

Ultimately, such advancements could empower planners to focus more on quality, design, and strategic thinking, rather than administrative tasks. While the potential of AI and digital technologies in planning is clear, the pace and complexity of implementation still face significant barriers.

Urban Planning: Slow Uptake and Structural Hurdles 

Despite the growing hype AI's integration into urban development remains limited and often misunderstood. Many confuse it with long-standing data modelling tools—such as those used in the planning of transport flows—but true AI, especially "strong AI", is far more sophisticated. The distinction between weak and strong AI matters, especially in policy discussions.

In London and the UK, AI is being used sparingly, often focused on relatively simple descriptions of land identification and its uses. There are cases of innovative uses of new data collection and analysis tailored to specific client’s needs and specs (hence very costly). Currently, the most significant innovations in AI are to be found outside of the UK in the fields of transport, environmental health and urban management. These are particularly visible in Asian cities, most notably in China, Malaysia, and Saudi Arabia.

A Fragmented Digital Planning Landscape

Over the past few years, a digital planning ecosystem has started to form within UK government structures—from the Ministry of Housing, Communities and Local Government (MHCLG) to the Greater London Authority (GLA), local councils, and private consultancies. Yet, no large-scale initiative has shown that AI is fundamentally transforming planning practices.

While urban planners are increasingly open to using AI, they face persistent barriers: poor data quality, high implementation costs, limited technical skills, scalability issues, outdated business models, legal uncertainty, and deep concerns about ethics as well as public and political trust.

Government ambitions to automate aspects of planning are further constrained by regulations—such as the legal obligation to respond to 100% of public consultation comments—which limit the scope for AI intervention. Moreover, the success of AI tools is heavily dependent on the availability of reliable, structured data—something many local authorities currently lack.

Pilot projects—typically small, interdisciplinary, and publicly funded—have been the main testing ground for AI in planning. Yet, without sustained investment and structural change, these remain isolated experiments. Private firms are also cautious: consultancies are built on billable hours and human-led processes, so efficiency-driven AI solutions risk undermining traditional revenue streams.

Real Estate: AI’s Leading Edge

In contrast, the real estate industry is quickly capitalising on AI’s capabilities—particularly in land acquisition, asset management, and profit optimisation. Start-ups combining AI, data analytics, and political risk tools are emerging to serve large firms that rely heavily on land value and development returns.

AI is already reshaping multiple stages of the property business cycle:

• Site acquisition: AI can identify strategic plots using demographic data, transaction histories, and gentrification indices.
• Design and development: AI is influencing architectural planning and project feasibility.
• Construction: Developers are using drones and mapping tech to monitor site progress in real time. 
• Sustainability: AI supports environmental assessments and helps model energy impacts.
• Building management: Tools are being developed to monitor tenant behaviour, space use, and profitability.
• Negotiations and decision-making: AI helps clients understand political trends, predict planning decisions, and even profile stakeholders.

While promising, these tools remain expensive and accessible mainly to major players. But the pace of adoption is accelerating.

Resourcing, Legal and Ethical Challenges

Remain Despite these advances, the path to widespread AI adoption in urban planning is fraught with legal, ethical, and societal concerns. One key issue is liability: it remains unclear whether AI-generated decisions or data are insurable or enforceable in contracts. Public planners are also bound by strict ethical codes. The use of AI for behavioural profiling or public manipulation is broadly seen as unacceptable, particularly in democratic settings.

There’s also a broader political sensitivity. AI technologies are sometimes associated with authoritarian surveillance systems, which stirs public mistrust. Without robust frameworks to address these fears, governments risk backlash and resistance.

Finally, access to high-quality data remains a fundamental barrier. Collecting or scraping the necessary data is costly and labour-intensive, often making it unattainable for local authorities under significant financial strain.

A Transformative Future—with Huge Potential

The potential of AI to transform urban planning is undeniable. With the right safeguards, it can enhance efficiency, speed up decision-making, and help planners to create functioning, adaptable and resilient places of high quality. But this technological shift could also reinforce existing inequalities. If left unchecked, AI-led planning may serve private interests over public good, marginalise vulnerable communities, and deepen intersectional disparities.

To move forward, regulatory reform, legal clarity, and public reassurance are essential. The challenge lies in striking the right balance between innovation and inclusion—between AI’s promise and its perils, as well as finding the right governance to support trust.

The AI revolution in city-making has already begun. But whether it leads to a smarter, fairer urban future—or an exclusive, profit-driven one—remains to be seen. Significant political decisions, at national level, are needed to create the relevant triggers to accelerate the use of AI in planning and its support to the delivery of better housing and living conditions.

Authors:
Professor Lauren Andres, Director of Research at the Bartlett School of Planning, UCL
Professor Mike Raco, Head of School in the Bartlett School of Planning, UCL.