top of page

Building Better Cities Through Trustworthy Technologies

  • Writer: Brydon Wang
    Brydon Wang
  • 3 days ago
  • 13 min read

By Dr Brydon Wang


One of the unexpected things I learned from watching my children build Lego cities was how quickly questions of governance emerge. At first, the exercise seems straightforward. They begin with the roads, move onto the buildings before parks spring up in the spaces between, the entire city growing in a beautifully organic manner.


Then somebody proposes a new rule.


Perhaps a road can now cut through a park or that a building can be twice as tall as the others, looming over its corner of the city. Perhaps a dinosaur is suddenly permitted to live in the central business district, sending its occupants screaming in terror towards the shadow lands below the bed.


The resulting argument (alas, somewhat inevitable for us parents!) is rarely about the specific change being proposed but something more fundamental: who gets to decide how the city evolves?


As adults, we like to think our cities are governed differently. Yet many of the disputes that shape contemporary urban life are remarkably similar. Communities argue about housing density, renewable energy infrastructure, transport corridors, flood mitigation, data centres and surveillance technologies. In Brisbane, we are already seeing these patterns emerge around Olympic infrastructure planning. Beneath every one of these debates sits again that same question, just dressed in different clothes: who gets to determine how our shared environments change?


For much of human history, those decisions were made by planners, engineers, architects, politicians, policymakers, developers and communities. Increasingly, however, technological systems are becoming active participants in the process. Their promised seamless, frictionless decision-making process, able to be carried out at scale and at a fraction of the time makes for a compelling narrative of efficiency that human judgement, with all its foibles alone, just cannot match.


Thus, if twentieth-century urban governance focused on who should make decisions, twenty-first century urban governance is increasingly concerned with how automated intermediaries make decisions on our behalf.

 

The Promise and the Problem of Optimisation

AI is now used to assist planning assessments, with algorithms influencing insurance pricing and the allocation of resources, infrastructure and services. Sensors collect real-time information about how cities operate, that then allow digital twins to be built on a constant stream of real-time data to simulate future development scenarios and automate decisions about how we plan, design, build, operate and manage our cities.

 

In my edited collection Automating Cities, published by Springer in 2021, I explored the implications of this shift. The question was never really whether cities would become more technologically sophisticated, because that seemed very much inevitable. The question I was more concerned with was what happens when decision-making within the city itself becomes increasingly automated.


At first glance, automation appears attractive because cities are extraordinarily complex systems. They generate vast quantities of information and involve countless decisions about land use, transport, energy, water, housing and public services. Automation, coupled with the new oil of data, promises increased visibility of conditions on the ground that can then lead to speedy, consistent decisions that can be repeated at scale. Faced with competing demands and limited resources, it is easy to understand why governments and institutions are drawn towards technologies that claim to optimise outcomes.


But optimisation is a deceptively simple idea and quite the trap.


Optimisation looks to refining a system or process in a way to deliver the best possible outcome. Implicit in that is the requirement to articulate an objective, that is, somebody must decide what ‘best’ means. This was one of the recurring themes in my research on smart cities, and what I later described as the seductive smart city. The promise of technology is typically framed in terms of efficiency: better data, better predictions, better decisions = better cities. And that formula is so seductive because it often stops us from asking the other question that matter… where does all this efficiency lead us to?


A system can optimise traffic flow while diminishing walkability or optimise energy distribution while overlooking localised impacts. It can optimise planning approvals while reducing opportunities for community participation. When we start making decisions only at a city scale and work to sublimate individual concerns into a series of numbers that we try to refashion into an ideal number… we run the risk of achieving a technical outcome that while may be impressive on the screen, may yet be soundly rejected by the communities the process is meant to serve. And with good reason, because the objectives being pursued were never agreed upon by those very same communities. The failure is not that technology fails to meet the expectations set but that the very expectations that have driven the design were never clearly tested, and in more cases than not, shared transparently.

 

The Right to Participate in City-Making

Henri Lefebvre argued in his 1968 book Le Droit à la Ville for what he called the ‘right to the city.’ The phrase is frequently interpreted as a claim about access to urban space, but Lefebvre's argument was considerably more ambitious. The right to the city was ultimately a right to participate in the production of urban life itself.


Cities are not natural phenomena. They are continuously created and recreated through decisions about land use, infrastructure, transport, public space, housing and investment flows. Consequently, where a community member exercises their right to the city, they participate in the processes through which those decisions are made in the three-way dance between public, private and civic spheres.


The significance of Lefebvre’s work is that it reframes urban governance as a question of participation rather than merely administration. Researchers including those at the Media Architecture Biennale have explored how digital technologies can create new opportunities for participation in city-making, recognising that technology is not merely something imposed upon cities but also a conduit through which community members can contribute to shaping a city’s future.


However, the mediation of screens and algorithms in an increasingly automated decision-making environment introduces a new and serious challenge. If city-making increasingly occurs through algorithms, optimisation systems, digital twins and automated decision-making tools, then participation becomes more complicated.


First, access to technology and how data points are collected need to be tested. The Sidewalk Labs project in Toronto faced intense criticism for attempting to privatise public areas and force community members and visitors to hand over personal data simply to navigate the neighbourhood. Critics and local technology leaders condemned the propose as an exercise in surveillance capitalism, warning against corporate take-over of traditional local government structures and even broader regulatory frameworks touching on privacy.


This pervasive monitoring plan relied on a vast network of sensor tracking everything from foot traffic to energy use. The resulting backlash led to high-profile resignation so privacy experts and legal action from the civic sector. Ultimately, public resistance and economic uncertainty forced the project to be abandoned during the pandemic in 2020.


Second, decisions that were once visible and contestable may become embedded within technical systems that are opaque, difficult to interrogate and resistant to challenge. The challenge is no longer simply who gets to participate in city-making. It is whether participation remains meaningful at all when decision-making becomes automated.


Third, one of the risks with contemporary discussions about democratic cities is that participation is often reduced to consultation, where community members are invited to comment on plans that have already been largely determined. Lefebvre's original formulation was far more radical. The right to the city was not merely a right to be heard but a genuine right to participate in the ongoing production of urban life. That distinction becomes increasingly important as cities become datafied.

 

The Problem of the Black Box City

For Lefebvre, urban space was not simply a physical container within which life occurred. Urban space was socially produced through countless interactions between institutions, infrastructure, economics, politics and everyday lived experience. Cities were always being remade through decisions about what should be (re)built or preserved, brought into existing networks or excluded.


As mentioned above, many of these decisions that support city-making are increasingly informed by data and driven by algorithms. However, in my research on digital twins, a specific risk emerged: that the seamless movement from data collection to automated decision-making can produce a black box city. Data is extracted from the built environment and its members, fed into automated decision-making frameworks, and translated into outputs that govern how people then live, move and work in the city, all with limited transparency about how those decisions were made or why they were made and on whose behalf.


This is a critical line of inquiry as urban life is increasingly shaped by automated decision-making systems. Participation in city-making may no longer depend solely upon access to political institutions but on access to the assumptions, objectives and optimisation criteria embedded within technical systems. If those systems are opaque, the right to the city becomes correspondingly hollow.

 

Vulnerability, Intermediaries and the Law

The law has a long-standing way of addressing this vulnerability that arises from information asymmetry that underpins the power imbalance between parties. Where information is unevenly distributed (ie. where one party knows significantly more than another), one of the options the law has taken is to contractually create a trusted intermediary.


In my earlier research on blockchain technologies (2016) and smart construction contracts (2018), I explored how many contractual functions could become increasingly automated. Much of the discussion at the time focused on efficiency: if contract administration could be automated, disputes reduced and payments accelerated, significant productivity gains could potentially follow.


But one of the most interesting questions concerned the role of the superintendent: the person appointed to administer construction contracts, certify payments, assess claims and make decisions affecting multiple parties. Although the superintendent's decisions are guided by contractual frameworks, they characteristically involve judgment, interpretation and the balancing of competing interests. The superintendent exists precisely because construction projects create significant information asymmetries where no single participant possesses complete information, and each relies upon representations made by others. The law does not create the superintendent because humans are inefficient but because the parties are vulnerable. As such, this legal response in the creation of trusted intermediary is not simply administrative.


In my subsequent research on the automation of construction contracts, what emerged was a fascinating question on how the automation and augmentation of the intermediary functions through sensors, machine learning systems, common data environments and smart contracts is that there is a corresponding attempt to automate what is ‘trusted’ about the trusted intermediary.


Advocates of blockchain and smart contracts often described these systems as ‘trustless’ because they reduce reliance upon central authorities. This description was always misleading as trust itself does not disappear in such systems but redistributed across nodes. Instead, it is less trusting of centralised authority, using a community-in-network that stands as trusted intermediary. But even such technologies still have trust bottlenecks in the form of the information oracle that pushes real-world (off-chain) information on-chain. And as much as the trusted intermediary may be replaced by a network, an algorithm, a data source, at its core design development, human intervention (again a centralised authority) still determines what and how information enters the system and the manner in which decisions are triggered / automated. The trusted human intermediary does not vanish, it is just embedded in code and replaced by a less visible one.


The same dynamic is increasingly visible in city-making.


Urban technologies are often presented as neutral tools. However, they perform functions that were once carried out by planners, engineers, administrators and policymakers. They classify, prioritise, recommend and allocate. In doing so, they become new forms of intermediary between communities and the decisions that affect them. This transformation creates intermediaries that are frequently less visible and far harder to hold to account than their human predecessors.

 

Trustworthiness as a Framework for City-Making

This is why I have been deeply interested in trustworthiness as a framework for evaluating how cities should be built and governed. In earlier works, I argued that discussions about artificial intelligence often focus on trust when they should focus on trustworthiness. The same observation applies to cities.


My research has drawn heavily upon the work of Mayer, Davis and Schoorman, who identified three key dimensions of trustworthiness: ability, integrity and benevolence. These three dimensions provide a useful framework for evaluating how urban technologies should be designed and deployed.


  • Ability signals convey the technical functionality of a system and how it can meet the claims of its developers. This can be signalled through accuracy of modelling, data reliability and currency, sound analytical frameworks and appropriate deployment. And it is clear that many of the data-focused technologies involved in city-making exhibit ability and augment human capability in genuinely useful ways, particularly in relation to quantitative assessments that would otherwise require significant human effort and be prone to human error.


  • Integrity signals are met where the outputs of these technologies enable decision-makers to produce results that meet community expectations (‘value congruence’). At minimum, these urban technologies need to be designed and have outputs that support decision-making that complies with regulatory frameworks, industry standards and community norms.


  • Benevolence signals demonstrate a genuine positive orientation towards those affected by decisions at an individual level. This is the most demanding dimension. A technically competent system can still produce outcomes within the strict letter of the law that communities eventually reject  even if these communities were part of the original law-making process. To address this dimension, technology developers need to create consensus building mechanisms that validate an assumed set of values that could be incorrectly signalled by laws and industry standards that may not meet that specific instance of technology deployment. Benevolence requires decision-makers to demonstrate that they understand and acknowledge the vulnerabilities of those affected by their decisions for each individual deployment scenario of the technology. It shouldn’t be a move fast, break things approach but one that is carefully calibrated to the needs of the community the technology is intended to serve.


Many contemporary debates about urban technology focus on ability. They ask questions about whether a model can predict flooding, or whether this digital twin can improve planning outcomes or optimise traffic flows. And while these are important questions, they are not the first questions that need to be asked.

Instead, the primary question that policymakers and technology developers need to land is whether the system has been designed with a positive orientation towards the interests of those affected by it. In other words: is it benevolent?

 

Transparency as Benevolent City-Making

In practical terms, benevolence often appears through transparency practices. But transparency is frequently discussed as if it were merely a technical feature, such as   an obligation to publish data or disclose methodology. I think we need a new approach to transparency as a governance mechanism.


Transparency practices should be oriented on the ‘how’: how it creates opportunities for communities to understand how decisions are being made; how assumptions are embedded into the system and how outputs are destabilised where these assumptions are proven to be wrong. Transparency practices need to invite scrutiny to create a forum for disagreement and contestability. Most importantly, it interrupts the seamless movement from data collection to automated decision-making that produces the black box city.


This argument was central to my article on the seductive smart city. The promise of technology is often framed in terms of seamlessness: frictionless systems that make better decisions faster. But democratic cities have always relied upon a certain amount of productive friction, such as debate, disagreement and participation. These are essential features of legitimate city-making and fundamental to democratic societies. The danger of treating these features as inefficiencies to be engineered away risks silencing communities and their right to the city.


The rapid growth of data centres provides a useful contemporary example. At one level, data centres are pieces of infrastructure. They consume land, electricity and water, but also create jobs and economic activity, providing support for the digital services that most of us use every day. However, they also raise deeper questions about allocation of resources and public interest, questions that have now become central to the social licence framing that is shaping data centre development globally and now emerging with greater urgency across Australia.


These are not purely technical questions but questions about governance. And as decision-making in these areas becomes increasingly informed by data, algorithms and automated systems, they also become questions about trustworthiness.

 

What Trustworthy Urban Technology Looks Like

If automation is displacing or augmenting traditional intermediaries in our cities, we need to consider how these new systems can be designed to demonstrate trustworthiness across all three signals, instead of concentrating on ability.

Several principles follow from this framework.


  1. The design of automated urban systems should make explicit the objectives being optimised and the values being prioritised. Optimisation is never neutral and policymakers and technology developers should make the set of objectives visible as the first step in making it contestable.


  2. These systems should be designed with what I have described in my digital twins research as seams: hold points in automated decision-making processes that create space for human intervention, review and the exercise of judgement. Seamless automation may be efficient but it does not meet the signalling needs for a technology system to be seen as trustworthy.


  3. The communities affected by these systems should have genuine access not merely to the generated outputs of the system but to the assumptions embedded within them. Transparency must extend beyond the decision to the decision-making process itself.


  4. Most importantly, the design of urban technologies should be oriented towards the vulnerabilities of those affected by them. Cities are, at their core, collections of shared vulnerabilities. Residents depend upon infrastructure they do not control and communities rely upon decisions that are often made for them. And when these systems fail due to floods, housing shortages, blackouts and infrastructure bottlenecks, these challenges reveal the extent to which urban life depends upon complex networks of interdependence... because technology does not remove these vulnerabilities, but merely redistributes them. As such, the signal of benevolence necessitates policymakers and technology developers asking whether that redistribution of vulnerability is being conducted with a genuine positive orientation towards those who bear the greatest risk.


Technology will undoubtedly play a central role in the future of cities. Whether those cities are worthy of trust will depend on how we choose to govern the decisions made to make these cities… and whether those making those decisions retain a genuine positive orientation towards the people who must live with the consequences. In this sense, trustworthiness is not simply a characteristic of technology but a form of urban infrastructure. Consequently, the challenge of building better cities is both a challenge of innovation and a continuous balancing exercise to ensure that the technologies shaping urban life remain accountable to the communities they affect through the trustworthy signals of ability, integrity and benevolence.

 

 Related research


 
 

© Brydon Timothy Wang

bottom of page