Experts Caution That AI-Driven Agility May Paralysis Britain’s Planning System | Planning Policy

The government’s initiative to leverage artificial intelligence for accelerating home planning could face an unforeseen hurdle: the agility of AI.

A new platform named Opponent is providing “policy-backed appeals in minutes” for those dissatisfied with nearby development plans.

Utilizing generative AI, the service examines planning applications, evaluates grounds for objections, and categorizes the potential impact as ‘high’, ‘medium’, or ‘low’. It also automatically generates challenge letters, AI-enhanced speeches for planning commissions, and even AI-produced videos aimed at persuading legislators.

Kent residents Hannah and Paul George developed this tool after their lengthy opposition to a proposed mosque near their residence, estimating they invested hundreds of hours in the planning process.

They’re making this service available for £45, specifically targeting people without the financial means to hire specialized lawyers to navigate the complexities of planning law. They believe this initiative will “empower everyone, level the playing field, and enhance fairness in the process.”

Though we are a small company, we aim to make a significant impact. A similar offering, Planningobjection.com, markets a £99 AI-generated objection letter with the slogan ‘Stop complaining and take action’.

Additionally, community activists have encouraged their audience to utilize ChatGPT for drafting appeal letters. One activist described it as like having a lawyer “ready to plan.”

A prominent planning lawyer cautioned that such AI could potentially “boost agility,” yet widespread adoption might overwhelm the planning systems and inundate planners with requests.

Sebastian Charles from Aardvark Planning Law noted that in their practice, no AI-generated objections contained references to prior litigation or appeal decisions, which were verified by human lawyers.

“The risk lies in decisions being based on flawed information,” he remarked. “Elected officials could mistakenly trust AI-generated planning speeches, even when rife with inaccuracies about case law and regulations.”

Hannah George, co-founder of Objector, refuted claims that the platform promotes nimbyism.

“It’s simply about making the planning system more equitable,” she explained. “Currently, our experience suggests that it’s far from fair. With the government’s ‘build, produce, build’ approach, we only see things heading in one direction.”

Objector acknowledged the potential for AI-generated inaccuracies, stating that using multiple AI models and comparing their outputs mitigates the risk of “hallucinations” (where AI generates falsehoods).

The current Objector platform is oriented towards small-scale planning applications, like repurposing an office building extension or modifications to a neighbor’s home. George mentioned that they are developing features to address larger projects, such as residential developments on greenbelt land.

The Labor government is advocating for AI as part of the solution to the current planning gridlock. Recently, they introduced a tool named extract, which aims to expedite the planning process and assist the government in fulfilling its goal of constructing 1.5 million new homes.

However, an impending AI “arms race” may be on the horizon, warned John Myers, director of the Inbee Alliance, a campaign advocating for more housing with community backing.

“This will intensify opposition to planning applications and lead to people unearthing vague objections they hadn’t previously discovered,” he stated.

Myers suggested a new dynamic could emerge where “one faction employs AI to expedite the process, while the opposing faction utilizes AI to impede it.” “As long as we lack a method to progress with desirable development, this stalemate will persist.”

Governments might already possess AI systems capable of managing the rising number of dissenting voices spawned by AI. Recently, they unveiled a tool named consult, which examines public consultation responses.

This initiative hopes to ensure “large-scale language models will see widespread implementation,” akin to those utilized by Objector, although it may merely increase the volume of consultation responses.

Paul Smith, managing director of Strategic Land Group, reported this month a rise in AI use among those opposing planning applications.

“AI-based opposition undermines the very rationale of public consultation,” he expressed in Building magazine. “It’s claimed that local communities are best suited to understand their areas…hence, we seek their input.”

“However, if residents opt to reject the system and discover reasons prior to submitting their applications, what’s the purpose of soliciting their opinions in the first place?”