When Data Becomes Knowledge: Governing Generative AI for Public Value in Cities

The concept of the smart city has been centered around sensors, cameras, and dashboards for decades. Today, conversation has shifted due to the rise of generative artificial intelligence (AI). Collecting data is no longer the sole objective; the focus has shifted to generating knowledge, automating intricate tasks, and revolutionizing the interaction between governments and residents.

In this new landscape, the central question is not technological but strategic: How do we build a strong AI governance framework that ensures artificial intelligence creates public value rather than deepening inequality?

 

Image Licensed via Adobe Stock.

 

What Generative AI Actually Brings to Cities
Cities today operate in an environment of growing complexity, where decisions must be made faster, with more data, and under greater public scrutiny. Generative AI strengthens urban capacity by automating routine tasks, translating complexity into clarity, and making technical processes accessible to a wider range of stakeholders. Its contribution can be understood through four core technical capabilities:

• • Creativity — generating novel ideas, designs, and alternatives that expand the range of options available to planners and policymakers.

• • Synthesis — integrating diverse datasets (spatial, administrative, environmental, behavioral) into coherent insights.

• • Prediction — forecasting outcomes based on historical patterns and real time signals.

• • Simulation — modeling complex urban scenarios so cities can test policies, infrastructure changes, or emergency responses before implementation.

Together, these capabilities allow cities not only to analyze the present but to anticipate and prototype long term solutions.

 

Why This Moment Matters for Cities
Cities face a convergence of pressures that make generative AI not just useful but necessary. Rapid population growth, resource constraints, and increasingly interdependent infrastructures have pushed urban systems to a level of complexity that traditional planning tools can no longer manage alone. At the same time, residents expect services that respond in real time to changing conditions, personal needs, and unexpected disruptions. In this context, generative AI becomes more than an automation tool: it acts as a strategic technological partner that augments human judgment, accelerates analysis, and supports more integrated and anticipatory decision making across departments.

Generative AI strengthens the ability of public institutions to govern complex issues with greater precision, transparency, and responsiveness by enabling cities to analyze large datasets, anticipate future scenarios, and translate technical information into clear, actionable insights. Its capabilities are rooted in deep learning foundation models, particularly Large Language Models (LLMs) which learn patterns at scale, interpret context, and generate original, coherent outputs that support informed decision making across urban systems.

Generative AI strengthens multiple dimensions of urban systems by translating its core technical capabilities into practical tools that support planning, operations, and public-facing services. Across these domains, AI functions as an enabling layer that enhances institutional capacity and broadens access to data-driven decision-making.

• • Urban Planning
    Rapid generation of zoning alternatives, infrastructure layouts, and development scenarios.

• • Citizen Services
    Personalized assistance, multilingual communication, and accessible information for diverse populations.

• • Governance
    AI lowers technical barriers by turning complex datasets into clear narratives and visualizations, enabling smaller municipalities and community groups to participate more fully in decision-making.

• • Emergency Response
    Scenario modeling, evacuation routing, and crisis communication.

• • Environment
    Climate impact assessments, pollution monitoring, and sustainability planning.

• • Energy
    Grid optimization, renewable integration, and consumption forecasting.

As these capabilities expand across urban systems, the question is no longer whether generative AI can support cities, but how institutions can guide their use toward public value. Technical potential alone is not enough; it must be anchored in governance frameworks that ensure transparency, equity, and collective benefit. This is where the work of building an AI governance model becomes essential.

 

Building a Governance Framework for AI in Urban Life
The growing integration of generative AI into urban systems positions it not as a technological accessory, but as a governance challenge and a territorial opportunity. For cities, AI intersects directly with questions of livability, proximity, and territorial equity. Recent global debates, including those emerging from the 2025 United Nations General Assembly, underscore that AI governance has become an urgent societal priority, requiring institutions to anticipate risks, align standards, and ensure that technological innovation serves the public good.
To harness AI responsibly, cities need governance frameworks grounded in four core principles:

1. 1. AI systems should operate transparently.

1. 1. Have equity safeguards to prevent biases.

1. 1. Ensure data stewardship that protects residents and provides public value.

1. 1. Be subject to democratic oversight to maintain accountability and human control.

Without these guardrails, AI can introduce significant risks, especially when adopted without regulation, purpose, or community involvement.

• • Algorithmic Bias → Territorial Inequality
    AI systems that learn from biased or incomplete datasets can perpetuate and intensify existing inequities. Critical domains like policing, credit scoring, mobility planning, and service allocation are affected by this. Algorithmic bias can exacerbate hidden territorial injustices in cities, perpetuating existing patterns of exclusion that are challenging to overcome.

• • Data Extraction and Privacy
    Rather than being responsible for their data, cities are in danger of simply providing it. Municipalities may lose control and public value when they rely on AI systems that require massive amounts of local data, as the power balance tilts in favor of external platforms. It is essential to protect privacy and ensure that data is used for collective good.

The increasing complexity of AI systems can render their decision-making processes incomprehensible, or ‘black boxes.’ As a result, residents are unable to understand, challenge, or dispute the decisions that impact their daily lives, since the outputs are unclear. The weakening of public accountability erodes trust in institutions.

 

The Missing Piece: A Governance Model for AI in Cities
The integration of AI in urban systems has sparked a crucial discussion, highlighting the need for a social agreement that guarantees the ethical, transparent, and socially beneficial use of AI in cities. The foundation of this pact is built on four essential pillars:

• • Transparency
    Understanding how models work, what data they use, and how they shape public decisions.

• • Equity
    Evaluating distributional impacts: who benefits, who is harmed, and which territories risk being left behind.

• • Public Value
    Ensuring AI strengthens inclusion, efficiency, and quality of life — not merely cost reduction.

• • Democratic Oversight
    Giving residents a voice in defining uses, limits, and priorities.

When these principles guide implementation, AI can be understood as part of a layered intelligence system within the city, combining perceptual (sensing), cognitive (reasoning), and generative (planning) capacities. Together, these layers allow cities to observe real‑time conditions, interpret complex patterns, and generate future scenarios through tools such as digital twins, simulations, and predictive models.

Fundamentally, the argument is that smart cities are not focused on implementing AI, but rather on regulating its use. This protocol is naturally aligned with the layered intelligence model; it makes urban intelligence transparent and easy to understand through clear outputs, promotes fair planning by identifying vulnerable areas, builds municipal capacity in cities with limited resources, and fosters participatory governance through interactive simulations and storytelling that citizens can understand. By integrating AI into public institutions, it prevents the concentration of power and avoids relying on private platforms for intelligence.

In this perspective, AI should be viewed as a shared cognitive infrastructure, rather than an external addition, and its governance must be a collective effort to promote livability, proximity, and territorial equity.

 

Towards Cities We Want
With generative AI increasingly integrated into daily urban life, cities must make a crucial decision. Alternatively, they can either adopt AI as separate tools or establish it as a unified cognitive infrastructure, which would enhance institutions, increase public value, and foster more livable and equitable territories. The key difference is not the technology, but rather the guiding frameworks, principles, and collective commitments that shape its utilization.

The purpose of AI, according to a human-centered approach, is not to replace human judgment, automate governance, or prioritize efficiency above all else. The city aims to improve its care for residents by making services more accessible, making decisions more transparent, and opportunities more evenly distributed throughout the city. With intentional governance, AI can become a powerful ally in promoting proximity, territorial equity, and dignity in everyday life.

Additional readings and references that support this article.

Hossam El Shoukry, Generative AI Opportunities in Smart Cities. Masterclass. ZIGURAT Institute of Technology, June 2025.

Beliz, G., IA: Hacia un nuevo pacto social tecnológica. Clase abierta. Escuela de Gobierno. Universidad Austral, April 2026.

United Nations General Assembly. (2025). Resolution 79/325: Terms of Reference and Modalities for the Establishment and Functioning of the Independent International Scientific Panel on Artificial Intelligence and the Global Dialogue on AI Governance. In Resolutions and Decisions Adopted by the General Assembly during its 80th Session (A/80/49, Vol. I). United Nations.