The bilent Battle for control of the city: AI is already governing

In 2025, the promise of “smart cities” has moved beyond aspirational rhetoric and into a field of direct economic and political competition. Artificial intelligence (AI) is no longer a technological add-on; it has become the new urban operating system. It determines which potholes are fixed first, how public transportation routes are adjusted in real time, where security resources are deployed, how water is prioritized during shortages, and which permits are accelerated to attract investment. The real debate is no longer whether AI will reach municipalities, but who controls it, under what rules, and whether its benefits will remain public or be captured by vendors, bias, and opacity.

The progress made in 2024 delivered an uncomfortable lesson for Mexico: AI adoption in the productive fabric remains marginal. The 2024 Economic Censuses show that only 26,093 business units reported using AI systems, roughly 0.5% of the total. This figure reveals a structural gap: if the private sector is adopting AI slowly, local governments—which depend on vendors, talent, and data—face even greater friction. At the same time, Mexico’s national statistics office reported signs of digitalization advancing, but at an insufficient pace: between 2018 and 2023, the share of establishments selling online rose from 1.6% to 2.8%. A smart city requires more than connectivity; it needs institutional capacity to procure technology strategically, govern it transparently, and audit it rigorously.

In the United States, 2024 revealed the opposite side of the same dilemma: rapid adoption paired with institutional anxiety. A survey cited by StateScoop reported that 48% of state and local agencies were using AI tools daily, compared to 64% at the federal level. The number is striking, but it also exposes a risk: moving fast does not necessarily mean moving well. At the municipal level, another warning sign emerged from IT leadership: 38% believed their local government was “not prepared at all” to use AI safely, with concerns centered on privacy, security, and a lack of skills. In short, the challenge has shifted from “what can AI do?” to “how do we prevent AI from eroding public trust?”

That said, AI-driven urban applications are already too valuable to ignore. In mobility, predictive models optimize traffic signals, reduce congestion, and prioritize public transit. In water management, AI detects leaks, estimates neighborhood-level demand, and enables predictive maintenance for aging infrastructure—an especially critical issue for Mexican cities facing water stress. In public safety and civil protection, algorithms combine reports, weather data, crowd levels, and risk signals to anticipate incidents, though here the red line is clear: without rules, algorithmic surveillance can amplify discrimination. In public administration, AI automates digital service desks, classifies requests, and shortens permitting timelines, directly impacting investment climates and regional competitiveness. The economic value is evident: the city becomes a platform for efficiency.

Yet 2025 added a geopolitical and binational dimension: technological sovereignty. In Mexico, the narrative has shifted toward infrastructure, data centers, and domestic capabilities; initiatives presented in 2025 around a “national language model” and the expansion of data centers show that AI is now understood as an industrial policy lever. In the United States, the pendulum has swung toward risk management and governance frameworks: the National Institute of Standards and Technology (NIST) continues to position guidelines for AI risk measurement and management, which are increasingly relevant for public procurement and system evaluation. From academia and national policy organizations, there is growing insistence that state and local governments adopt frameworks such as the NIST AI Risk Management Framework and municipal toolkits to avoid improvisation.

With these signals, 2025 is shaping up to be the year of “rules and trust.” New York City, for example, has advanced guidance and public debate around government use of generative AI, aiming to align innovation with public-sector values. This type of governance foreshadows what more cities will adopt: procurement contracts with audit clauses, algorithm registries, impact assessments, and minimum standards of explainability.

In my view, the greatest challenge in 2025 will not be technological, but institutional and political. First, data: without interoperability, quality, and cybersecurity, AI merely automates disorder. Second, talent: municipalities compete with the private sector for professionals who understand data, digital law, and urban operations. Third, public procurement: if contracts fail to require metrics, traceability, and audit rights, cities become captive to vendors.

Fourth, democratic legitimacy: if residents perceive “black boxes” making decisions, the smart city project quickly turns into a crisis of trust. Fifth, the binational gap: while the U.S. normalizes usage and debates regulation, Mexico must accelerate local capacity so AI is not just imported technology, but a driver of productivity, public services, and measurable regional development.

AI is already rewriting urban life. For Mexico and the United States, the question is no longer whether the city will be smart, but whether it will be fair, secure, and governable.

Written by: Editorial

#SmartCities #ArtificialIntelligence #DigitalGovernment #MexicoUS #PublicInnovation