First Published:
17 Jun 2026, 12:10 pm
Tags:
Details
Reviewed by
Ola Söderström, Sophia Maalsen, Mark Whitehead, Ayona Datta, Jennie Day, Federico Cugurullo, Federico Caprotti, Matthew Cook, Andrew Karvonen, Pauline McGuirk and Simon Marvin
Book review forum: Artificial Intelligence and the City: Urbanistic Perspectives on AI
Cugurullo Federico, Caprotti Federico, Cook Matthew, Karvonen Andrew, McGuirk Pauline, and Marvin Simon, Artificial Intelligence and the City: Urbanistic Perspectives on AI, London and New York: Routledge, 2023; 420 pp.: ISBN: 9781003365877.
Introduction
Ola Söderström – University of Neuchâtel, Switzerland
In the time between the publication of Artificial Intelligence and the City and this review forum, the presence of AI in our lives and in the public sphere has intensified at a breakneck pace. This acceleration is the very logic driving generative AI companies as they engage in a frantic race to raise funds, multiply their data centres, increase their computing power, and develop new user interfaces. We are already a long way from the short-lived call issued in 2023 by more than 1000 technology leaders (including Elon Musk) for a pause in AI development. The race and fierce competition among major AI companies are now unchecked in a context where any form of government regulation tends to be portrayed as a competitive disadvantage. Contrary to previous calls for caution and ethical safeguards, American leaders in the AI industry have opportunistically embraced the authoritarian shift in U.S. politics. Thus, politically, the current transition toward an AI society (Törnberg and Uitermark, 2026) is taking place within the context of the rise of techno-authoritarian libertarianism (Slobodian, 2025). To cite just one example, as I write this, U.S. and Israeli forces are bombing Iran and Lebanon, relying on artificial intelligence to identify targets and locations and compress decision time.
Given this techno-political acceleration, it might seem that Artificial Intelligence and the City is already outdated. This is not the case. The strength of this volume lies in its proposal of a historically grounded conceptual framework. It allows readers to move beyond singular case studies and to understand the specificities of urban AI through its historical trajectory. As the book’s editors point out, AI is characterized by its autonomy and agency, as in the case of autonomous lethal weapons used on battlefields, as well as autonomous vehicles operating on public roads. It is also characterized by its tangible presence in our daily lives that has become very evident due to the massive energy and water demands of data centres (Cugurullo et al., 2025), in contrast with the more invisible technologies of the smart city.
The comments gathered in this forum therefore all rightly highlight the importance of this book, which helps to characterize the co-constitutive relationship between AI and the urban world. These comments also suggest interesting avenues for further research and reflection. Sophia Maalsen suggests exploring further the intersections between gender and technology, as well as our cyber-symbiotic relationships with AI. Mark Whitehead questions the divide between AI and the smart city posited by the book’s editors and invites us to consider what is truly artificial about AI. Ayona Datta, for her part, invites us to further analyse the role of the state (particularly at the local level), data-poor cities, and gig labour in the deployment of AI urbanism. Finally, Jennie Day suggests we continue exploring socio-technical imaginaries of AI, as well as the values and beliefs of AI engineers. The stimulating future developments discussed in this review forum will be able to build on the milestone represented by this important edited collection.
References
Cugurullo F, Caprotti F, Day J, et al. (2025) The nature of AI: Metabolism, energy, water, labour and justice in the urban political ecology of artificial intelligence. Urban Political Ecology 1(1–2): 33–54.
Slobodian Q (2025) Hayek’s Bastards: The Neoliberal Roots of the Populist Right. Penguin Random House.
Törnberg P, Uitermark J (2026) From the platform society to the AI society: Towards critical studies of generative AI. Available at: https://osf.io/preprints/socarxiv/qahd3_v1 (accessed 3 April 2026).
Commentary 1
Sophia Maalsen – University of Sydney, Australia
Each year I give a lecture to students in the Masters of Urbanism on the different phases of technology adoption in cities. I start with the cybernetic cities of the 1960s and 1970s (Project Cybersyn is always a hit!), before moving onto the computational cities of the 1980s and onwards. Next are the smart cities of the late 2000s and 2010s which still persist despite discussions around platform urbanism and now, AI urbanism. Distinguishing these last three—smart, platform, AI—is not always straightforward and is something which prompts questions from students. To what degree is AI urbanism different from smart urbanism? This question is not just asked in the classroom: an audience member at AAG 2025 also posed the question during a session on AI urbanism. I don’t always feel confident in my answer to such questions, but Artificial Intelligence and the City (Cugurullo et al., 2023) is a good place to start.
AI and the City does well to define and sketch out the emergence of urban AI as an area of geographical research. Particularly useful is the acknowledgement of multiple and heterogeneous urban artificial intelligences, which reflects the multiple aspects of the urban to which AI is being applied. This is not a book then that gives us one narrow working definition of urban AI. Rather the book usefully identifies the core characteristics of AI urbanism and illustrates these with examples across four types of urban AI used to organize the contributions into four sections: autonomous vehicles, urban robots, city brains, and urban software agents.
One of the contributions of the book is the work it does to draw out distinctions between smart cities and urban AI. There has been much discussion around whether the adoption of AI in cities is distinct from the preceding smart cities of the last 15 years, and this book adopts the stance that we are indeed witnessing something new in the long history of technology use in cities. The shift from smart cities to AI urbanism is, according to this work, represented by a shift from technologies being used to count and calculate urban phenomena to technologies being used to produce an account of urban phenomena by collecting and analyzing data to explain how and why certain things occur in cities. The collected chapters all contribute insights on this shift in some way.
While this shift is seen as transcending smart urbanism, from what I am seeing in my own work, many of the lessons of the smart city have not been learnt in this transition to AI urbanism. It makes me question, then, whether we should not be so quick to distinguish the two but rather approach them along a continuum and continue to ask questions as to why the critiques and faults of the smart city are continuing to play out. Much like the smart city, discourse around AI position it as a seamless solution to managing cities—something which smart tech was unable to fulfil once it encountered the messiness of urban life on the ground. This will likely be a challenge for AI as well. For example, in my current work with Robyn Dowling, Pauline McGuirk, and Claire Daniel, we are looking at AI in planning trials in NSW, Australia. Our initial interviews with planners are showing a mix of hope and skepticism that AI will be able to deal with nuances, complexities, and geographical variations inherent in planning, and indeed the early feedback has been that this has been a challenge.
One of the key insights from this book that is helping our current thinking is around the use of experimentation and pilots of AI, whether AVs or robots, as having a desensitizing effect to the extent that AI urbanism is seen as inevitable. Trialing in low-risk scenarios and applying AI to mundane tasks risks desensitizing critique and prefigures an AI future that limits resistance or allows for alternative paths. As such, I have found AI and the City useful for my thinking around what AI urbanism is, the challenges it presents, and the critiques we need to sustain around its adoption in cities.
There are, however, some areas which I think could be further developed. The first is gender, which remains relatively unproblematized throughout, apart from an occasional reference to the feminine nature of digital assistants, Jackman’s (2023) chapter on drones, and Lin and Yeo’s (2023) chapter on airport robots. There is a long tradition of feminist critique of technology that draws attention to how encounters with technology are gendered and I find it limiting that these debates are not engaged with in more depth, especially considering the urban itself is experienced differently dependent on a number of intersectional positions. Although there is an understanding of urban AI as multiple and heterogeneous, I think greater attention could have been paid to the multiple experiences of AI based on difference. This would be an important area for future work. The second is exploring in more detail the ways people “work with” AI in cities. The contributions do exemplary work showing how despite promises of autonomy, much of the application of AI still relies on human labor as well as needing to interface with humans. This is a great start and as the field develops further, I would like to see more questions asked around these interactions. For example, in my thinking on algorithms as collaborators, I ponder what would be the epistemological, methodological, and empirical opportunities provided by acknowledging the work that AI does as we work with it (Maalsen, 2023). An entry point into this is thinking about the type of relationships we build with AI as we do this work of city making. The concept of cybersymbiosis (Hayles, 2021; Maalsen, 2024) is helpful here. Building on Haraway’s (2016) idea of oddkin, we can approach AI as our “digital oddkin” (Maalsen, 2024). For Haraway, making kin is a political act with a commitment to making better worlds and allowing for multispecies to flourish. How could we make a better urban future for our cities by working with our digital oddkin? What would a cybersymbiotic city look like?
These are the questions I am asking, and I have no doubt that this collection will continue to be a reference point as I ask these questions. As it stands, next year, when I give my lecture on technology and cities, I will be referring to this book to help myself and my students think through the “so, what is different” questions that are necessary to ask.
References
Cugurullo F, Caprotti F, Cook M, et al. (eds) (2023) Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge.
Haraway D (2016) Staying with the Trouble: Making Kin in the Chthulucene. Duke University Press.
Hayles NK (2021) Three species challenges: Toward a general ecology of cognitive assemblages. In: Lindberg S, Roine HR (eds) The Ethos of Digital Environments: Technology, Literary Theory and Philosophy, 1st edn. Routledge, pp.27–45.
Jackman A (2023) Everyday droning: Uneven experiences of drone-enabled AI urbanism. In: Cugurullo F, Caprotti F, Cook M., et al. (eds) Artificial Intelligence and the City. Routledge, pp.114–135.
Lin W, Yeo SJI (2023) Airport robots: Automation, everyday life and the futures of urbanism. In: Cugurullo F, Caprotti F, Cook M., et al. (eds) Artificial Intelligence and the City. Routledge, pp.169–185.
Maalsen S (2023) Algorithmic epistemologies and methodologies: Algorithmic harm, algorithmic care and situated algorithmic knowledges. Progress in Human Geography 47(2): 197–214.
Crossref Web of Science Google Scholar
Maalsen S (2024) Home in cybersymbiosis: Making home with digital oddkin. Information, Communication & Society 27(9): 1759–1775.
Crossref Web of Science Google Scholar
Commentary 2
Mark Whitehead – Aberystwyth University, UK
When it comes to artificial intelligence, William Gibbons’ observation that “the future is already here. It is just not very evenly distributed” seems particularly apt. I say this because Artificial Intelligence and the City (Cugurullo et al., 2023) takes us as close as it seems possible to currently get to empirically understanding the elusive impacts of AI on everyday urban life. The book transports us into the urban heartlands of an AI future that is far from evenly developed, allowing us to discern what this near future may mean for us all. Delving into AI’s application in autonomous vehicles, urban robots (including household drones), urban governance platforms, digital brains, and the delivery of urban health and social care, this volume reveals the highly uneven—and often messy and problematic—deployment of AI in urban spaces. What emerges is a form of critical travelogue, traversing the nascent spaces of AI experimentation and transient public laboratories in cities like London, New York, Singapore, and Hangzhou.
As far as I am aware, this is the first book-length volume to offer a holistic, critical analysis of the impacts of AI on urban life and planning. Its 20 chapters provide an impressively diverse and comprehensive account of the complexities of urbanism and AI. The volume is both empirically rich and theoretically pluralistic. It wisely avoids offering a synoptic analysis of AI urbanism, instead serving as a foundational text for critical inquiry. In what follows, I offer three provocations—essentially incitements for debate on how best to construct critical geographies of the technological present and future, which emerge directly from the chapters of this volume.
Artificial Intelligence and the City opens with the statement, “Innovation in artificial intelligence (AI) is transforming cities in unprecedented ways” (p. 1). This justificatory context is regularly used to preface work on AI in both the social and computational sciences. To this extent, it feels like a statement of the obvious. What is interesting, and commendable, about this book is that it devotes significant analytical energy to undermining the very premise on which it is based. The volume highlights AI’s unprecedented ability to spread automated decision making throughout urban life. We all know the argument: AI can digitally perceive and analyze the complexities of urban space in ways that human cognition—often biased and bounded—cannot. However, across various chapters, we see that when it comes to urban AI, humans are rarely “out of the loop.” The very idea of autonomous, AI-based decision making is often a necessary illusion: an attempt to generate a veneer of faux-automation, behind which the commercial value of AI can be burnished.
In the case of automated vehicles, for instance, the book explores how AI presents an unprecedented challenge to the automobile-centric design of cities, but ultimately reveals AI systems that primarily aim to improve the comfort and convenience of car and truck driving. Other chapters consider AI’s potential to alleviate fiscal struggles in cities and enhance public services through technological solutionism. Yet the volume also helps expose the ways AI is used to obscure the neoliberal erosion of social care and public services in urban spaces.
In these contradictory contexts, the volume invites us to question the extent to which AI is actually transforming cities and to reconsider the implications of framing this through the hyperbole of the “unprecedented.” I was left wondering whether a future research focus should center on what transformations AI might be inhibiting; in other words, positioning AI as a wholly precedented—and even counter-revolutionary—force.
A key justification for this volume is the assertion that AI requires analytical sensibilities that transcend current work on smart technology. It was this post-smart trajectory that I found most problematic. At one level, I am sympathetic to the editors’ intention of contextualizing the book in this way. The higher-order computational functionality of AI urbanism, and its promise for automated decision making, might suggest a clear distinction from the more modest prospects of smart cities. However, the assertion of a dawning post-smart era felt premature and too easily won. The subsequent chapters often returned to familiar, if still valuable, accounts of smart technology. This should perhaps not be such a surprise—urban AI is still very much facilitated by smart tech. The capacities of AI would not be possible without the digital capture and coding of public, private, and secret lives already achieved by smart technologies.
Interestingly, this volume reveals that the emergence of AI itself is partly driven by the real-world impacts of smart technology. In a chapter exploring autonomous lorries, for instance, we learn that AI in trucking is being necessitated, to some extent, by the cognitive overburdening of drivers caused by the saturation of their cabs with various forms of smart technology. In this context, what the volume helpfully reveals—though it may not intend to—is the path dependencies and deep interdependencies of smart technology and AI. While I understand the value of emphasizing the novelty of AI urbanism and the importance of thinking afresh about the role of ever-more intelligent digital technology in urban spaces, I think assertions of a post-smart era may lead us to too hastily abandon the hard-fought critical insights we have developed about intelligent technology in our headlong pursuit of technological novelty.
The book introduces some interesting perspectives on how we might think about “urban” AI, and the role that geography and urban studies play in sensitizing us to the spatial dynamics and implications of artificial intelligence. The volume also offers nuanced reflections on the nature of digital “intelligence” and its distinction from biological cognition. However, what is missing in the volume—and indeed in much computational and social scientific analysis of AI—is a deeper exploration of what precisely makes AI “artificial.”
For me, the framing of intelligence as “artificial” within discussions of AI carries significant moral implications, which are often overlooked. The term “artificial” suggests a separation from the biological (human) element of cognition. While AI seeks to mimic human intelligence, it implies a denial of the human within it: AI looks, sounds, and functions like human intelligence, but is often presented as functionally distinct—and superior. This moral framing of AI serves to return the human to a state of “first nature,” implying the futility of human regulation of a technology we barely understand.
At times, this volume reflects some of the challenges of exploring the complex relationships between AI and urban life. While humans are sometimes foregrounded in discussions, in other chapters their presence is conspicuously absent. Ultimately, the book led this reader to muse on the value of future research that is framed less exclusively around urban artificial intelligence and instead considers the coming together of multiple biological/ecological and non-biological intelligences within cities.
I firmly believe that Artificial Intelligence and the City will become a staple reference for work in the field of Urban AI. It offers perceptive perspectives on our collective assumptions about digital transformations. It also foregrounds how these assumptions shape our analytical projects and their ability to offer critical insights. At this point in the evolution of AI-enabled cities, this volume is a crucial contribution to thought and inquiry.
References
Cugurullo F, Caprotti F, Cook M, et al. (eds) (2023) Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge.
Commentary 3
Ayona Datta – University College London, UK
This ambitious book is one of the first in this field to bring together a range of case studies examining the stickiness of artificial intelligences from across several parts of the global north and south. The key argument in this book is that AI is urban not by chance but by design. The calculability and predictability of AI futures thrive on data concentrations that are most evident in urban centers. In examining the significance and reach of AI urbanism, Cugurullo et al. (2023) argue that artificial intelligences particularly thrive in the city due the critical mass of data available to harvest and operationalize “pluriversal AI futures.” In doing so, Cugurullo et al. (2023) have created what they call a “post-smart cities” agenda for urban studies by conducting a “state-of-the-art review of AI urbanism from its historical roots to its contemporary global significance.” They conceptualize AI urbanism as extending along three main axes—function, presence, and agency. AIs function as essentially data-gathering machines, and establish their enduring presence in all aspects of urban life, and this gives them a characteristic agency of acting without human presence albeit within parameters originally defined by humans. Examining a range of AI-driven technologies across four sections on autonomous vehicles, urban robots, city brains, and urban software agents, they suggest that despite the emergence of AIs as aspirational urban futures, their logics of emergence are situated in the simultaneous history of urban and technological developments, the boundlessness of the extractive nature of technologies, and the possibilities of multiple paradoxical spaces to coexist and overlap across digital and geographical boundaries.
I was reading AI and the City while I was deep in fieldwork in 2024, examining the complexities and nuances of state-led land digitalization projects in India, Kenya, and Mexico. Inspired by this book, I asked the Commissioner of a small municipality in India how technology can contribute to urban governance in his city. The Commissioner passionately described that his aspiration was to use AI robots to do the sanitation work in his city and thereby eliminate caste-based discrimination in manual scavenging. This aspiration in a city which faces major challenges in providing urban basic services to much of its population represents the fantasies of an automated ubiquitous vision of “digitalising states” (Datta, 2023). AIs nonetheless have real-life impacts on the cities and urban spaces where they exist as a discursive fantasy, channeling valuable resources and state capacity away from poorer neighborhoods toward the pursuit of AIs. As Amoore (2020: 5) notes, it is the way in which “algorithms are implicated in new regimes of verification, new forms of identifying a wrong or of truth telling in the world” that makes their presence heightened in such ethico-political urban domains.
While this book is a solid accomplishment on charting an agenda on AI urbanism, there are however three aspects of AI urbanism that I would wish to see further research on in the future. First, the role of the state in conducting the “future craft” of AI is often underexplored. The deployment of this fantasy by the state to transform cities is often an opportunistic power brokering in the present. In the shift from smart cities to AI governance, the role of the state has become even more relevant in producing a mirage of urban certainty vested in AI. As Crawford (2021: 186) notes, “The state is taking on the armature of a machine because the machines have already taken on the roles and register of the state.” Given the visible legacy of rudimentary and frugal infrastructures (Guma, 2020) co-existing alongside fantasies of AI urbanisms, we need to pay particular attention to the relationship between AI and the local/municipal state in urban regions.
Second, there is more research needed on how AI urbanism might fail in data-scarce cities and city regions. Beyond the metropolises in the global south, AI as an aspirational discourse is derailed by the lack of data to train algorithms to learn. This is particularly relevant in the case of land digitalization platforms that seek to increase revenue bases of municipalities but are built upon incomplete and missing land information databases (Demerutis et al., 2025; Hoefsloot and Gateri, 2024). Similarly, autonomous vehicles in cities with dense traffic and/or broken road infrastructures are bound to fail. On the other hand, AI-based facial recognition systems trained on poor data will generate false positives in urban surveillance. Indeed, most cities in the global south will only be able to experiment with fragmented technologies of AI urbanism, with immense challenges in scaling up across all cities.
Finally, more attention needs to be drawn to the question of gig labor as the foundation of “artificial” intelligences. As Crawford notes, “AI is neither artificial nor intelligent. Rather, AI is both embodied and material, made from natural resources, fuel, human labor, infrastructures, logistics, histories, and trajectories” (Crawford, 2021: 125). AI unfolds because of the enormous array of digital workers who are often precarious, casualized, and located within a global gig economy, and who conduct the work of tagging, annotating, and training the machine. Indeed, human labor is the cognitive intelligence of a computational machine.
To understand the significance and impacts of AI urbanism in the future, we need to look further into the processes and steps toward algorithmic futures—digitization, digital augmentation, digitalization, and full automation—particularly following the trajectories toward full, partial, and fragmented AI futures across the global north and south. This book cannot answer all these questions. But it has created a solid foundation from which to explore the ethical and moral landscapes of AI urbanism, asking why certain forms of AI futures are possible while others are not, and why AI unfolds so ubiquitously in some regions and not in others.
References
Amoore L (2020) Cloud Ethics: Algorithms and the Attributes of Ourselves and Others. Duke University Press.
Crawford K (2021) Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press.
Cugurullo F, Caprotti F, Cook M, et al. (eds) (2023) Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge.
Datta A (2023) Digitalising state: Governing digitalisation-as-urbanisation in the global south. Progress in Human Geography 47(1): 141–159. https://doi.org/10.1177/03091325221141798.
Demerutis JA, Datta A, De Jesus Flores J (2025) Informational peripheries by design: Land, layer, and leverage in metropolitan area of Guadalajara (MAG). In: Datta A, Hoefsloot FI. (eds) Informational Peripheries: Rethinking the Urban in a Digital Age. UCL Press, pp. 67–90.
Guma PK (2020) Incompleteness of urban infrastructures in transition: Scenarios from the mobile age in Nairobi. Social Studies of Science 50(5): 728–750. https://doi.org/10.1177/0306312720927088.
Hoefsloot FI, Gateri C (2024) Contestation, negotiation, and experimentation: The liminality of land administration platforms in Kenya. Environment and Planning D: Society and Space 42(5–6): 645–663. https://doi.org/10.1177/02637758241254943.
Commentary 4
Jennie Day – University of Ottawa, Canada
Artificial Intelligence and the City (Cugurullo et al., 2023) interrogates the logics of urban AI by carefully charting the historical antecedence of, and mutually constitutive relationships between, AI and urban society. Motivated by the desire to expose “the depth of the changes driven by urban AI and the need to understand them before they become irreversible” (p. 365), the book forcefully asserts that AI urbanism represents a significant political, economic, and ideological departure from smart cities. Artificial Intelligence and the City empirically and theoretically examines the rise of AI urbanism through diverse case studies ranging across autonomous mobility, drones, governance platforms, real estate, social care assistance, policing, healthcare, urban design, and beyond.
While considerable scholarship in urban and digital geography examines the labor conditions of those subjected to urban AI platforms, AI and the City charts a novel path by considering the “upstream” labor of the workforces designing and developing urban AI. Excellent urban scholarship unpacks the precarity of the labor force whose “platform-mediated labor” is at the mercy of extractive “hyperlean digital platforms” (Moore and Bissell, 2023: 206) such as the on-demand gig economy services of Uber, Deliveroo, and the like (Altenried, 2021). AI and the City, however, makes a valuable contribution to urban studies by examining the labor of the individuals who develop the AI technologies and thereby actively bring urban AI futures into being. For instance, in Chapter 15, Shapiro (2023) examines how software developers building predictive policing software make sense of the law enforcement context. The chapter interrogates the developers’ technosolutionist perspectives which understand their urban AI products as essential “solutions” to identified policing “problems.” Relatedly, in Chapter 13, Adam Moore and David Bissell analyze the technical and “more-than-technical” labor required to bring about “platform cooperativism.” These are alternative forms of urban AI platforms that push back against the “hegemony of privately owned, data-hungry, and hyper-capitalist platforms” (Moore and Bissell, 2023: 207) and are instead both owned and operated by the individuals who use them. Both pieces provide incisive insights into the labor required to develop urban AI, although their analysis remains close to the technology being examined—whether it be developers’ thoughts about how policing impacts law enforcement technologies or the multiple forms of labor required to develop alternative platforms.
Responding to the editors’ call for further examination of imaginaries, I suggest that fruitful exploration of urban AI politics from a labor perspective can interrogate future imaginaries and broader worldviews of the AI engineers designing, developing, and undertaking hegemonic urban AI future making. Imaginaries are referenced in AI and the City in a broad range of thematic and empirical contexts: driverless freight imaginaries (p. 60); spatial imaginaries (p. 105); seductive, benign imaginaries (p. 105); personal and cultural imaginaries (p. 157); geographical imaginaries (p. 101) and beyond. Notably, in Chapter 14, Chen (2023) foregrounds imaginaries when examining the performed imaginaries of AI-controlled cities in China. Drawing on Jasanoff’s (2015) notion of sociotechnical imaginaries, Chen assesses the gulf between the rhetoric of nation state AI imaginaries and how these imagined AI futures are brought into being in the actually existing city of Xiong’an.
Future imaginaries direct our attention to the “collectively held, institutionally stabilised and publicly performed visions of desirable futures” (Jasanoff, 2015: 322)—focusing particularly on the politics of claims made in the present about the future. The notion of future imaginaries facilitates the investigation of which actors are given the authority to stake a claim to the future—thereby unevenly facilitating certain futures and foreclosing others (Halford and Southerton, 2024; Wajcman, 2017). Future representations, discourses, and imaginaries have the capacity to be constitutive and generative of futures as they can assemble likeminded stakeholders, direct agendas, and shape the investment of resources in the present (Beckert, 2016).
AI professionals inherently transmit their personal values into the technology they construct (Haider, 2017) and, consequently, designing and developing AI technologies is a profoundly value-laden project (Wu, 2025). Therefore, a fruitful way to interrogate the politics of urban AI futures is to unearth the beliefs, perspectives, and worldviews that are held by those developing the technology. Following English-Lueck (2017), who asked Silicon Valley tech workers to discuss how they envisaged the future in order to uncover their implicit assumptions and values, undertaking a critical interrogation of the urban AI future imaginaries held by those individuals actively creating urban AI futures will enable an exploration of how “imagination and intention shapes choices and plans” (English-Lueck, 2017: 18). In practice, this entails uncovering and analyzing the implicit assumptions, values, priorities, ontologies, worldviews, and priorities inherent in urban AI engineers’ visions of what constitutes a good life, a good society, a good city, and a healthy relationship between technology and urban life.
This approach productively extends those undertaken in AI and the City by going beyond a limited focus on the imaginaries of a given urban AI technology at hand. It actively assesses AI innovators’ wider beliefs about science and technology, what comprises just and equitable sociodigital futures, together with “how the future is mobilized in real time to marshal resources, coordinate activities and manage uncertainty” (Brown and Michael, 2003: 4). Such an exploration both provides a framework for critically exploring the future politics of urban AI and, in practical terms, enhances AI engineers’ ethical imaginary by providing a portal into the political consequences of the future making they are actively undertaking.
References
Altenried M (2021) Mobile workers, contingent labour: Migration, the gig economy and the multiplication of labour. Environment and Planning A: Economy and Space 56(4): 1113–1128.
Crossref Web of Science Google Scholar
Beckert J (2016) Imagined Futures: Fictional Expectations and Capitalist Dynamics. Harvard University Press.
Brown N, Michael M (2003) A sociology of expectations: Retrospecting prospects and pospecting retrospects. Technology Analysis & Strategic Management 15(1): 3–18. https://doi.org/10.1080/0953732032000046024.
Chen B (2023) Performed imaginaries of the AI-controlled city: Conducting urban AI experimentation in China. In: Cugurullo F, Caprotti F, Cook M., et al. (eds) Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge, pp.223–239.
Cugurullo F, Caprotti F, Cook M, et al. (eds) (2023) Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge.
English-Lueck JA (2017) Cultures@SiliconValley. Stanford University Press.
Haider S (2017) The darkness at the end of the tunnel: Artificial intelligence and neoreaction, Viewpoint Magazine. Available at: https://viewpointmag.com/2017/03/28/the-darkness-at-the-end-of-the-tunnel-artificial-intelligence-and-neoreaction/ (accessed 3 April 2026).
Jasanoff S (2015) Future imperfect: Science, technology, and the imaginations of modernity. In: Jasanoff S, Kim S-H (eds) Dreamscapes of Modernity: Sociotechnical Imaginaries and the Fabrication of Power. University of Chicago Press, pp.1–33.
Halford S, Southerton D (2024) Sociodigital futures? An agenda for sociological research and practice. Sociologia Italiana 25(3): 93–110.
Moore A, Bissell D (2023) Encountering limits in cooperative platforms: The more-than-technical labour of urban AI. In: Cugurullo F, Caprotti F, Cook M., et al. (eds) Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge, pp.205–222.
Shapiro A (2023) Optimizing the immeasurable: On the techno-ethical limits of predictive policing. In: Cugurullo F, Caprotti F, Cook M., et al. (eds) Artificial Intelligence and the City: Urbanistic Perspectives on AI. Routledge, pp.240–257.
Wajcman J (2017) Automation: Is it really different this time? The British Journal of Sociology 68(1): 119–127.
Crossref PubMed Web of Science Google Scholar
Wu Y-H (2025) Capturing the unobservable in AI development: Proposal to account for AI developer practices with ethnographic audit trails (EATs). AI and Ethics 5(6): 5705–5718.
Authors’ response
Federico Cugurullo1,2, Federico Caprotti3, Matthew Cook4, Andrew Karvonen5, Pauline McGuirk6, and Simon Marvin7
1Department of Geography, National University of Singapore, Singapore
2Department of Geography, Trinity College Dublin, Ireland
3Department of Geography, University of Exeter, UK
4School of Engineering and Innovation, Open University, UK
5Department of Architecture and Built Environment, Lund University, Sweden
6School of Geography and Sustainable Communities, University of Wollongong, Australia
7The Urban Institute, University of Sheffield, UK
When we started writing this book, AI was beginning to emerge mostly through urban experiments that were reminiscent more of science fiction than real-life urbanism. Since then, the presence and perception of AI have rapidly evolved, and now when we talk about AI urbanism it is not in relation to alternative distant futures. Rather, we often have in mind something that we are experiencing in the cities where we live.
In our initial exploration of AI urbanism, we noticed strong connections with established practices of smart urbanism, and we agree with the reviewers’ comments on the smart–AI continuum which, as we illustrate in the book, is part of long-standing techno-urban transformations (Karvonen et al., 2019; Marvin et al., 2015). We are not arguing that AI urbanism constitutes a “clean break” or radical rupture with the smart city, even less some sort of reaction to it. Yet it is clear to us that the application of AI in cities involves significant points of departure from smart urbanism, as we also discussed in our recent papers (Caprotti et al., 2024; Cugurullo et al., 2024). For example, from a conceptual point of view, how can the notion of smart cities, grounded in ideals of modernity whereby technology is traditionally seen as an instrument firmly controlled by humans, explain the emergence of agentic urban AIs, such as LLMs, whose inner mechanics transcend human control? How can we understand (and indeed regulate) technological systems that are progressing from the boundedness of automation to AI-mediated spaces that are continuously unfolding, learning, and adapting? And to what extent can theories about the real-time city, with its focus on optimizing present conditions, help us understand urban AI’s fixation on anticipating the future (Cugurullo et al., 2024; Kitchin, 2014)? From our perspective, the literature on smart cities provides a useful starting point to interrogate the multiple implications of AI urbanism. At the same time, there is a need for urban socio-technical theory to develop new ways to grapple with the unique aspects of urban AI.
We also recognize that AI urbanism unfolds unevenly, thriving in data-rich metropolitan contexts while producing partial, experimental, or fragmented formations elsewhere. In these urban contexts, humans and the unique configurations of socio-political urban processes play a central role in the emergence of AI urbanism. After all, the diffusion of urban AI is embedded in socially produced ideologies and political economies (Cugurullo, 2025). Urban AI should therefore be understood less as a technological break and more as an intensification and reformatting of existing socio-technical assemblages. Overall, it is precisely the nexus of humans and AI in the contemporary city that gives us hope for more sustainable futures since, unlike machines, we can leverage alternative future imaginaries as stepping stones toward more humane uses of AI in our cities.
ORCID iDs
Ola Söderström https://orcid.org/0000-0002-7111-8765
Sophia Maalsen https://orcid.org/0000-0001-6384-0785
Mark Whitehead https://orcid.org/0000-0001-6499-4719
Ayona Datta https://orcid.org/0000-0002-0360-5406
Jennie Day https://orcid.org/0000-0001-6851-1999
Federico Cugurullo https://orcid.org/0000-0002-0625-8868
Federico Caprotti https://orcid.org/0000-0002-5280-1016
Matthew Cook https://orcid.org/0000-0003-2373-2127
Andrew Karvonen https://orcid.org/0000-0002-0688-9547
Pauline McGuirk https://orcid.org/0000-0002-6688-9661
Simon Marvin https://orcid.org/0000-0001-5538-5102
References
Caprotti F, Cugurullo F, Cook M, et al. (2024) Why does urban Artificial Intelligence (AI) matter for urban studies? Developing research directions in urban AI research. Urban Geography 45(5): 883–894.
Cugurullo F (2025) AIdeology: Unpacking the ideology of artificial intelligence and its spaces. Antipode 58(1): e70065. https://doi.org/10.1111/anti.70065.
Cugurullo F, Caprotti F, Cook M, et al. (2024) The rise of AI urbanism in post-smart cities: A critical commentary on urban artificial intelligence. Urban Studies 61(6): 1168–1182.
Crossref Web of Science Google Scholar
Karvonen A, Cugurullo F, Caprotti F (2019) Inside Smart Cities: Place, Politics and Urban Innovation. Routledge.
Kitchin R (2014) The real-time city? Big data and smart urbanism. GeoJournal 79(1): 1–14.
Crossref Web of Science Google Scholar
Marvin S, Luque-Ayala A, McFarlane C (eds) (2015) Smart Urbanism: Utopian Vision or False Dawn? Routledge.
