THE USE OF ARTIFICIAL

INTELLIGENCE IN CITIES
What opportunities and challenges does AI pose
in cities and how can international organisations
support AI’s ethical use?

A needs-based, mission-oriented and ethical use

of AI, including genAI, is the most responsible
way to create smart cities that deliver a
sustainable urban future and enhance the well-
being of their populations, leaving no one behind

MAY 2024
Andras Szorenyi
TABLE OF CONTENTS

INTRODUCTION 1

USE OF AI IN CITIES 1

ETHICAL USE OF AI BY LRGS 2

INTERNATIONAL ORGANISATIONS AND THE USE OF AI 4

GUIDING PRINCIPLES FOR USE OF AI BY LRGS 6

CONCLUSIONS 8
 adaption. Public institutions are not the
INTRODUCTION quickest to adapt, but local authorities are
It is predicted that artificial intelligence (AI) usually praised to be more reactive to meet
will generate great technological changes in the needs of their populations.
our lives. Alongside the Internet of Things
(IoT), blockchain, image processing and USE OF AI IN CITIES
digital apps, AI is considered to be among the
most disruptive technologies; it is already By 2050, about 70% of the world’s population
making impacts in numerous areas, including will live in cities. Cities attract people because
education, healthcare, business, agriculture, they offer a broad array of services and
and urban development.1 Each year opportunities - in education, health,
generative AI acquires new capabilities, employment, housing, water and sanitation,
becomes more accessible, and scales up, transportation, as well as cultural activities. If
creating massive opportunities and risks. cities can learn to use AI properly, they will
manage their increasingly complex socio-
It is widely assumed that AI and AI- economic systems more efficiently.
associated technologies will enhance
economic growth and public well-being in The ‘smart city’ is an urban development
‘smart cities’. This paper argues that local concept. Smart cities aim to improve their
and regional governments (LRGs) should residents’ quality of life by applying new
deploy these technologies, but only when technologies (AI, IoT, ICT, digitalisation, etc.)
they contribute to LRGs’ public mission and to make life more sustainable, resilient,
enhance the quality of life and meet the adaptive, and efficient. A smart city relies on
needs of their citizens. an ecosystem of objects and services that
interact with each other and with their users.
The paper looks at the opportunities and This being said, smartness is a tool, (use of
challenges that the use of AI might pose in technology) and not an aim in itself; the aim
cities. It reviews where AI can best serve being a city that is more sustainable, resilient,
local governments, and gives an overview of adaptive, efficient, etc.
the guiding principles that ensure AI’s ethical
use. It argues that LRGs should be able to Using AI technology, for example, cities can
feed their needs and expectations into global improve:3
processes for AI regulation. • Traffic management
• Public transport
“The last few years in AI have seen a shift
away from using multiple small models, each • Parking systems
trained to do different tasks—identifying • Waste management
images, drawing them, captioning them— • Infrastructure maintenance
toward single, monolithic models trained to
do all these things and more.”2 The business
sector has already taken important steps for

1
Ursino, D., de la Prieta Pintado, F., Artificial Intelligence as a Disruptive Technology—A Systematic Literature
Review.
2
Heaven, W. D., What’s next for AI in 2024 | MIT Technology Review.
3 Berry, I., 10 ways AI can be used in Smart Cities | AI Magazine.

1 I The use of Artifical Intelligence in Cities

In other areas, the use of AI is hotly debated people-centred approach. AI and other
due to the challenges it presents: technologies can serve a common purpose if
• Security and surveillance their implementation respects human rights
and foster sustainable development. To
• Energy consumption and distribution
balance technological opportunities and
• Predicting future needs socio-economic needs appropriately, and
• Measuring the environmental effects of ensure they are fully equipped to govern, to
AI infrastructure collect and analyse data, and to regulate,
Cities have also started to rely on ‘digital adopt and deploy AI-led solutions, local
twins’. This concept goes one step further, governments must invest in both building
offering a holistic model of sustainable urban their capacity and meeting international
planning. By combining digital technology, standards. International organisations can
including AI-enabled algorithms, with urban help them in this complex endeavor.
operational mechanisms, it offers the
opportunity to leap straight to a future-proof
urban upgrading. “Through the precise
mapping, virtual-real integration and
intelligent feedback of physical and digital
cities, it promotes safer, more efficient urban
activities and more convenient and inclusive
everyday services, as well as helping to
create more low-carbon, sustainable
environments.”4

Smart cities are highly exposed to the

disruptive nature of AI; indeed, they are the
outcome of disruptive innovations. However,
cities need to take care to ensure that
technological changes benefit all their
citizens. Cities that are smart and ETHICAL USE OF AI BY
sustainable, from online services to
developing citiverses, must have solid LRGs
foundations; they must be able to apply AI The massive use of AI in cities will probably
applications across urban management, land “usher in a period of hardship and economic
use, data analytics, business development pain for some whose jobs are directly
and environmental sustainability while using impacted and who find it hard to adapt —
data ethically and meeting the need for local what economists euphemistically call
public services. “adjustment costs.” However, the genie is not
AI and AI-enabled solutions provide new going back into the bottle. The forward march
opportunities for service provision and of technology will continue, and we must
support future-oriented urban development harness the new capabilities to benefit
but to do this successfully cities must adopt a

4
WEF, Digital Twin Cities: Framework and Global Practices | World Economic Forum (weforum.org).

2 I The use of Artifical Intelligence in Cities

society.”5 Regulated and standardised • Encourage local schools to teach AI
frameworks as well as associated and adopt AI and digitalisation.
procedures to integrate technology in • Invest in AI R&D and deploy AI
governance and decision-making, are applications innovatively and to good
required to make cities truly smart. effect.
• Encourage the public to discuss the
Understanding the value of risks and opportunities of AI to meet
local needs.
locally-generated data and
how data should be used for Data and data management are
the common good are the particularly important issues. The
primary responsibilities of adoption of AI implies the adoption of data-
local and regional driven decision-making, in various fields and
governments. at different levels of government. This
change impacts job markets, healthcare,
local businesses, and education. LRGs have
Obviously, operationalising new
a primary responsibility to understand the
technologies is also associated with
value of locally-generated data and how data
challenges and risks. These risks include
should be used for the common good. They
bias and discrimination, worsening
need to establish a robust framework to
inequality, and human rights violations. AI-
regulate national or local data, protect data
based solutions need to be lawful and ethical,
privacy, and address data concerns
and should involve all relevant stakeholders.
associated with AI.
“The private sector must lead the way in
funding and supporting a responsible AI “Data is the new oil – and it’s a good thing.”7
ecosystem. The solutions lie in open Cities generate hundreds of petabytes of
collaboration between policymakers data from which advanced machine-learning
establishing guidance frameworks, academic solutions can be derived. Data are sourced
communities developing safety mechanisms, from public service providers as well as
tech firms sustainably integrating AI, and civil private businesses. Local governments are in
society keeping its promise of empowerment a position – but sometimes lack the capacity
for all.”6 – to regulate the ownership and use of data
that they harvest from their own activities and
LRGs play a key role in developing the
by arrangement with private service
local AI ecosystem for smart cities. They:
providers.
• Build awareness at city level of AI’s
potential to solve local problems. The Global Cities Hub recommends that
• Developing procurement policies that LRGs should make sure that all the data
deploy AI ethically and incentivise its they collect and use serve a mission-
ethical use. oriented purpose and the common good.
• Encourage urban businesses to use AI.

5
Agrawal, A., Gans, J., Goldfarb, A., and ChatGPT, ChatGPT and How AI Disrupts Industries (hbr.org).
6
Minevich, M., The Dawn Of AI Disruption: How 2024 Marks A New Era In Innovation (forbes.com).
7
Bhageshpur, K., Data Is The New Oil -- And That's A Good Thing (forbes.com).

3 I The use of Artifical Intelligence in Cities

INTERNATIONAL International Geneva is at the forefront of the
‘AI for Good’ movement. Therefore, the
ORGANISATIONS AND THE Global Cities Hub recommends using the
USE OF AI city’s convening power – based on the
broad multistakeholder presence – to
Most international organisations have started discuss AI-related disruption, including
to integrate AI in their work, and some with LRGs, and develop common
support initiatives that seek to shape and narratives and definitions.
regulate AI at international level. This section
reviews multilateral processes, many of The International Telecommunication
which involve multistakeholder cooperation. Union (ITU) supports technical studies, and
It suggests that LRGs should be involved brings governments, industry and other
in international regulatory processes stakeholders together to improve the
from their inception. reliability, security, and interoperability of
urban ICT infrastructure. The ITU contributes
Various multilateral initiatives seek to to sustainable smart city development by
regulate AI. The UN Secretary-General’s setting standards and promoting global
High-Level Advisory Board on AI has been collaboration. Its ‘Guide for smart and
asked to map the governance landscape and sustainable city leaders’9 sets out a strategy
propose policy and institutional options in for sustainable local digital transformation.
preparation for the Summit of the Future in The AI for Good Global Summit is a leading
September 2024. The Council of Europe is United Nations platform that promotes the
currently negotiating an international use of AI to improve health, climate, gender
instrument on the development, design and rights, inclusive prosperity, and sustainable
application of AI systems. infrastructure; its concerns are highly
relevant to LRGs.10 ‘People-centered smart
In Geneva, which gathers an important
cities’ is a flagship programme of un-habitat.
number of international stakeholders, “most
It provides strategic and technical support on
organizations seem to work in their silos,
digital transformation to national, regional
which makes it difficult to speak with a strong
and local governments. The united nations
voice in global AI conversations. The majority
innovation technology accelerator for
of international institutions based in Geneva
cities (unitac) develops innovative solutions
have already embarked on the AI journey,
that accelerate achievement of the sdgs, and
either by applying AI tools internally or by
tests ideas, tools, scenarios and solutions
contributing to or building AI-enhanced
before deploying them in cities around the
instruments for global benefit. Directly and
world.11 the risk framework presented in un-
indirectly, Geneva is a part of how
habitat’s white paper ‘ai & cities’ (2022)
consequential AI decisions are taken.”8

8
Radu, R., The variable geometry of AI governance (genevapolicyoutlook.ch).
9
ITU, UNEP, WMO, UNU, Guide for smart and sustainable city leaders: Envisioning sustainable digital
transformation (itu.int).
10
ITU, AI for Good Global Summit 2024 (itu.int).
11
UN-Habitat and UNITAC, United Nations Innovation Technology Accelerator for Cities (UNITAC) Hamburg |.

4 I The use of Artifical Intelligence in Cities

promotes a people-centered approach to ai LOOKING BEYOND GENEVA HERE IS A
and community engagement.12 FURTHER LIST – JUST FOR FUN
GENERATED BY CHATGPT:
According to the High Commissioner for “While there might not be specific guidelines
Human Rights, Volker Türk, “The human exclusively tailored for the use of AI by local
rights framework provides the foundation we governments, several general AI ethics and
urgently need to innovate and harness the governance frameworks provide principles that
potential of AI”.13 The use of AI affects can be adapted and applied at the local level.
several human rights, including the rights to Here are some relevant guidelines and guiding
non-discrimination, to privacy, to work, and to principles for the use of AI:
an adequate standard of living. LRGs
consistently face such issues, which are 1. OECD Principles on Artificial
concentrated in the dense social Intelligence. Developed by the
environment of cities. OHCHR’s programme Organisation for Economic Co-operation
B-Tech,14 which seeks to improve the tech and Development, these principles provide
sector’s implementation of the Guiding guidance for responsible stewardship of
Principles on Business and Human Rights, is trustworthy AI. The principles cover
particularly relevant. Human rights fairness, transparency, accountability, and
inclusivity, and can be applied by LRGs.
mechanisms, notably UN Special
Rapporteurs, are producing reports on key 2. EU Ethics Guidelines for Trustworthy AI.
issues, including the right to privacy in the The European Commission's guidelines
emphasise human-centric AI that respects
digital age, and the impact of AI technologies
fundamental rights, ethical principles, and
on freedom of expression.15
societal values. Local governments can
The World Economic Forum (WEF) Centre align their AI strategies with these
guidelines.
for Urban Transformation supports
collaboration in cities to create more 3. IEEE’s Global Initiative on Ethics of
sustainable, inclusive, and resilient local Autonomous and Intelligent Systems.
IEEE's framework offers principles for the
communities and economies. WEF’s White
ethical design and development of AI
Paper ‘Governing Smart Cities. Use cases for
systems. It considers transparency,
urban transformation’ profiles governance accountability, privacy, and inclusivity.
and policy programmes designed to achieve
4. UNESCO Recommendation on the
responsible and ethical technology adoption.
Ethics of Artificial Intelligence. UNESCO's
It addresses three policy challenges: making recommendation emphasises that AI
data available to the public; ensuring privacy; should respect human rights, diversity, and
and improving access.16 the environment. Local governments can
use these principles to ensure they benefit
society while minimizing potential harms.
5. G20 AI Principles. The G20’s high-level
principles address transparency,

12
UN-Habitat, AI & Cities, https://unhabitat.org/ai-cities-risks-applications-and-governance.
13
OHCHR, Human rights must be at the core of generative AI technologies, says Türk | OHCHR.
14
OHCHR, B-Tech Project | OHCHR.
15
OHCHR, Is AI a force for good? | OHCHR.
16
WEF, Governing Smart Cities: Use Cases for Urban Transformation | World Economic Forum (weforum.org).

5 I The use of Artifical Intelligence in Cities

 accountability, fairness, and privacy. While GUIDING PRINCIPLES FOR
primarily aimed at national governments,
they are also relevant for the development USE OF AI BY LRGS
of AI policies at the local level.
When they adopt guidelines and principles
6. AI4People's Ethical Framework for a
for the efficient and ethical use of AI-enabled
Good AI Society. Developed by the
technologies at local level, LRGs should
AI4People initiative, this framework
proposes ethical guidelines for the
choose an inclusive multistakeholder
development and use of AI, including approach. International organisations should
principles such as transparency, also involve LRGs in their workstreams, to
accountability, fairness, and sustainability. help local decision-makers find effective
7. National AI Strategies. Some countries smart strategies for their citizens.
have developed national AI strategies that
The new ‘Guiding principles for artificial
include principles and guidelines for the
ethical and responsible use of AI. Local
intelligence in cities’, published by U4SSC
governments can refer to these strategies (ITU), sets out principles, enablers and
and adapt relevant principles to their governance methods for LRGs. It
specific contexts.” emphasises that ethical AI should be:

• Lawful;
• Privacy preserving;
• Fair and inclusive;
• Explainable and transparent;
• Accountable;
• Safe and secure;
• High performing and robust;
• Assessed for impact and sustainability;
• Enabling human autonomy.

Local governments have an increasing

number of AI based tools at their disposal.
The guiding principles can help them deploy
these responsibly and successfully.
International cooperation will continue to be
critical because “exchange of knowledge at
the local, regional, and international levels
will help develop AI principles formulation
and implementation and will also increase its
sustainability in the long run”.17

17
U4SSC, U4SSC Guiding principles for artificial intelligence in cities (itu.int).

6 I The use of Artifical Intelligence in Cities

UN-Habitat also supports national and “The guidelines, which will be developed by
local governments to harness digital 2025, will serve as a non-binding framework
technologies. It encourages the inclusion of for developing national and local smart city
marginalised and vulnerable groups in smart regulations, plans and strategies to ensure
city and digital transformation processes. UN that digital urban infrastructure and data
Habitat has been mandated to develop contribute to making cities and human
International Guidelines on people- settlements sustainable, inclusive,
centred smart cities (see t he table below). prosperous and respectful of human rights.”18

18
UN-Habitat, International guidelines on people-centred smart cities | UN-Habitat (unhabitat.org).

7 I The use of Artifical Intelligence in Cities

CONCLUSIONS and exploited ethically, to advance their
mission and the common good.
Smart cities that want to deliver better public
services and increase the well-being of their Various multilateral initiatives seek to
residents need to deploy AI and genAI in regulate AI and genAI. It would make sense
ways that meet needs and advance their to use Geneva’s convening power of a broad
mission. AI initiatives should align with multistakeholder presence to encourage
LRGs’ overall strategic priorities. LRGs that discussion of AI-related disruption, to involve
set clear goals and objectives can LRGs, and to develop common narratives
communicate their vision effectively and are and definitions.
in a position to make sound policy and
The Global Cities Hub recommends that
technological choices.
relevant international organisations
The various guidelines and guiding should formally and actively involve
principles developed or under LRGs in an institutionalized manner in AI-
development by relevant international related global decision-making
organisations serve as frameworks for processes, because cities are able to
creating national and local smart city significantly enhance the implementation
regulations, plans and strategies to ensure of international instruments.
that digital urban infrastructures and data
make cities and human settlements more
sustainable, inclusive, prosperous and
respectful of human rights.

Cities are especially active in innovation

using AI. Smart cities are able to control and
predict the consequences of planned
services and infrastructures. To profit from
the use of technology, LRGs need to set clear
standards for the ethical use of AI. They also
play a key role in developing local AI
ecosystems for businesses, and making sure
that new local technologies are used
efficiently and ethically. International
organisations should involve LRGs in their
efforts to promote viable, human-centred
smart cities.

Local governments are in a position to

regulate data ownership and use in their
own activities and by arrangement with
private service providers. LRGs should
ensure that all data is collected, processed

8 I The use of Artifical Intelligence in Cities

 THE GLOBAL CITIES HUB
The Global Cities Hub (GCH) connects cities, local and regional governments to the International
Geneva ecosystem.

Connecting all relevant stakeholders, the Global Cities Hub pursues 3 main objectives:

1. Facilitating the participation of cities and their networks in relevant multilateral

processes and bodies.
2. Facilitating partnerships between cities, their networks and Geneva-based
international organizations and other entities.
3. Urban mainstreaming by providing a space to discuss urban issues and enhance the
visibility of the urban work of Geneva-based actors.

The Global Cities Hub © The Geneva Academy of International

Villa Rigot Humanitarian Law and Human Rights
Avenue de la Paix 9,
1202 Geneva - Switzerland

Phone: +41 22 559 46 00

Email: [email protected]
www.globalcitieshub.org