Why every city needs to take action on AI

By Hannah Miller and Isak Nti Asare


When we think about the impact of artificial intelligence (AI) we usually ask one of several questions: what can it do? Where is it headed? And how widespread will its effects be? The growing consensus around these questions is something like: it can probably do less than you think right now, it may someday do more than you imagine, and it will therefore probably affect almost every area of your life. The rise of automation has given way to a fourth question: what should we do about it?

National governments, including France, China, and the United Kingdom, have already begun to create wide-ranging and forward-thinking policy documents to lay out strategies for maximising  AI’s benefits, and mitigating its potentially negative impacts. We recently published a blog reviewing existing national AI strategies, and the five key themes that national strategies share:

Diagram displaying the core components of national AI strategies

Diagram displaying the core components of national AI strategies

Amid the discussion about national governments creating strategies to prepare their countries for artificial intelligence, we think it is important that city governments also take time to develop a clear view on the impact of AI on their citizens. These should include the challenges and opportunities of automation specific to their economies and societies.  

It is easy in a national policy, for example, to note that AI will create more jobs than it replaces. But findings like this will vary markedly between cities. For example, a city with sizeable portions of the workforce employed in transport or logistics will be disproportionately affected by automation. A report from the Institute for Spatial Economic Analysis illustrates this point. They found that some cities in the United States, such as Las Vegas and El Paso, Texas, risk job losses of over 63 per cent by 2025. Another report by The Centre for Cities similarly found that places like Mansfield and Sunderland in the north of England are at risk of losing nearly 30 percent of jobs to automation. Taking action now at the city and regional levels can help to prepare workforces and welfare systems for the economy of the future.

Beyond the broader recommendations of a national strategy, a city strategy for AI will enable local leaders to focus on the specific challenges of automation for their area. It will help them to identify opportunities and prepare their workforces and communities to thrive in the fourth industrial revolution. Cities should be asking about the composition of their workforces, and specifically how they will be affected by automation. They should be developing options for re-training and upskilling the workforce, while looking to facilitate innovation via startup incubators and growth-hubs. More broadly, city governments that are involved in welfare provision should be re-examining their welfare systems and thinking about how they can harness automation to improve public services.

Obviously, all cities are not the same. The structural and institutional forms that shape cities vary from country to country: cities do not have the same policy tools or resources at their disposal. As a consequence, their responses to AI are going to be varied and that is exactly the point. City governments need to identify and employ every tool available to them to prepare for the risks and potential of artificial intelligence. Richard Schragger underscores this point in his book City Power where he examines the limits and potential of urban governance in the United States. Schragger argues that the power of cities is more relevant to citizen well-being than ever before. Despite the limits to city policy-making that leaders may face in their municipalities, Schragger urges city leaders to implement policies that improve basic service delivery while reducing the burdens of inequality.

As growing urbanisation coincides with the rise of automation, Schragger’s call to action resonates more clearly. Cities around the world have already taken initiative in fighting climate change, protecting the rights of migrants, and becoming hubs for innovation; Despite the variety of policy-making circumstances in the more than 4,000 cities with populations exceeding 100,000 people, it is now time for cities to lead in responding to AI as well.  

Lifelong learning initiatives such as the UNESCO Global Network of Learning Cities and the city-wide trials of Universal Basic Income taking place (in Barcelona; Stockton, California; and Hamilton, Ontario), are all examples of things that proactive cities are doing to prepare for automation. Crafting a comprehensive AI strategy at the city level will focus resources and efforts to harness the potential of AI while preparing citizens for the ways that automation will affect them.

AI has the potential to create seismic changes in the way we work and live. This is true at a national level, but the challenges and opportunities of AI will be experienced differently in cities. City governments should be anticipating these changes to ensure that they manage them properly. Forward-thinking cities should now demonstrate their leadership by creating and publicising AI strategies designed for their city, helping to ensure that their citizens are as prepared as they can be for the coming rise of automation.

We will be publishing a series of insights into the potential risks and pitfalls of AI for cities, and why cities need comprehensive AI strategies, over the coming weeks. In the meantime, we are working on designing effective AI policies for cities. If you would like us to help design a strategy for your city or local area, please get in touch with Isak at isak.ntiasare@oxfordinsights.com or Hannah at hannah.miller@oxfordinisghts.com

Hannah Miller