How cities are making data-driven decisions in the age of AI
Busakorn Pongparnit via Getty Images
By
Chris Teale
For almost a decade, What Works Cities has helped cities understand how to use the data they collect to improve outcomes. The advent of AI has accelerated much of that mission.
Cities have excelled at collecting data, but the recent growth of artificial intelligence has brought into focus the need for that data to be clean, standardized and ready to use.
Unclean data could lead to all kinds of negative consequences from AI tools, like incorrect outputs, hallucinations and bias, and those concerns have city leaders thinking hard about what to do next.
Thankfully, help is at hand. For almost a decade, Bloomberg Philanthropies has backed What Works Cities, a certification that recognizes local governments for their use of data to inform policy decisions and funding priorities while improving services.
More than 200 cities now have some level of certification, with more likely to follow as local leaders look to make sure their data is in a good condition and plan for an AI-driven future.
Route Fifty caught up with Rochelle Haynes, What Works Cities’ managing director, after a panel on cities’ use of AI at the recent Smart City Expo USA conference in New York. Among the topics of discussion were the future of AI, how to recognize when data is clean and what comes next for the organization.
This interview has been edited for brevity and clarity.
Route Fifty: We heard a few people speak yesterday about AI and cities. Where are we with this technology, particularly with its use in cities? What's your take on the landscape so far?
ROCHELLE HAYNES: Within the What Works Cities network, we have over 200 cities represented. Those cities range from cities that are just starting out building what we call their data infrastructure, to cities that are much more advanced with their data infrastructure. The same sort of continuum you see with that, you see with AI. We have cities that are cautious but optimistic. They're curious, they're approaching it from a place of wanting to understand it, wanting to establish some guidelines and governance around it, to cities that are already testing and using it and elevating use cases and giving their staff guidance to be able to use it. What we're seeing is cities want to be innovative in this period.
Local governments are mindful that often they're behind on adapting to technology, and they feel like this is a moment. What's been exciting in our network is elevating use cases of how AI has helped with operational efficiencies, how it's helping with things that feel very practical to city leaders. Whether it's flooding and being able to do predictive analytics, whether it's helping with the housing permit process in Burlington, Vermont, those are examples we like to uplift, because it gives folks something very tangible on how they can use AI.
Route Fifty: Is there anything that's really exciting you about AI's use in cities?
HAYNES: My favorite one is the chatbots. Two Argentinian cities have spearheaded this, both Buenos Aires as well as Corrientes. What I love about the chatbots is it's democratizing how people can access their government. I can access it 24/7 with my schedule, with my needs, and get my answers to my questions. It doesn't require someone to be there at midnight if that's when the emergency comes up or the question comes up. It's also allowing folks to be able to get nuts and bolts from cities faster. Permit processes are going faster, and you've seen a decrease in wait times. That's tangible to me, and that's exciting. It's getting folks able to connect with their government and have them be able to get the response that they need in a timely manner.
Route Fifty: Panelists talked a lot about making sure the data is clean. When is data clean enough? You can clean something and clean something, but when is it kind of clean enough to use?
HAYNES: A lot of cities have data in different places and different formats, and there may not be an owner of the data. When we think about cleaning data, are we collecting, storing and managing our data in the same system? Are we using the same fields? Are we using the same sort of nomenclatures? That, to me, is the clean data piece of it. It's never going to be perfect, especially when you're talking about cities, and especially with larger cities, but what we're trying to get to is really some level of standardization. That way you can start to use technology like AI. But perfection is not needed, not for our program, but neither for the use of AI. You need to have your data in an order so that you can tell a story from it and make some bit of sense of it.
Route Fifty: Does it need a mindset shift among some cities? They might think, if it's not perfect, we can't do it.
HAYNES: The conversations we have one on one when we start to meet cities, is letting them know we're here as a resource for you. You are not meant to be perfect for our resources. We want to help you advance in your data practices. We are a free resource that you can use to do that. One of the ways that we think about that is lowering the entry point for our program, and that's something that we're thinking about.
We don't want to be intimidating. We don't want to ask you too many questions before you start to access our capacity building workshops, or make you feel like we're only going to highlight the big cities and how big cities are doing great work, which they are, but I also want to highlight small cities that don't have as many resources, so that folks can start to see themselves in the work and see themselves in the examples that we're uplifting. It's a lot of conversations, but it's also, what is that front door you look at when you first learn about What Works Cities? We continue to think about how we can optimize that to not be intimidating.
It’s the same thing with AI. I think these conversations need to be simplified so that folks can feel like it's something that they can access. Especially the past year or so, it's a hot topic and it continues to be a hot topic. We'll continue to be talking about this, especially in these spaces, for some time. But it almost feels like you go from the leap of, “We don't have any AI,” to robots, and that's a huge leap mentally for people to make.
You’ve got to meet people where they are, and you’ve got to meet local governments where they're at. Where they're at is, can this help with a daily task that we have? You heard Mayor Andre Sayegh from Patterson, New Jersey, say yesterday on the panel that he’s resource strapped as a city. Do you have an AI tool that I can now deploy that's going to meet the needs of the staff member that I can't hire? That's tangible. For me, that's how we can start to make this conversation feel a bit more approachable. AI can be an added resource capacity building support that makes your job efficient and saves you time.
Route Fifty: I want to go back to what you said about lowering the entry point to get involved in What Works Cities. How has the organization evolved over the time that you've been here? Has the mission shifted at all?
HAYNES: The idea was, how do we build local capacity to address challenges: it's climate change, it's homelessness, it's failing infrastructure. As we've evolved, the program has started to say: we still have that as a mission, but we've now gone global. We're now not just a U.S.-based program. We're not just a domestic standard of excellence. We're an international standard of excellence on what it means to be good, well managed local government. We want to elevate use cases and best practices globally, because all cities are dealing with housing issues and climate change, youth unemployment. Everyone is grappling with it. The program has evolved in that regard, in that we're global now.
The program has also evolved in the sense that our rigor is the same. The certification standard isn't changing, but we are going to change what it means to access our services. We're going to be changing our assessment. Right now, there's about 43 different questions that we ask cities to answer before they access our services. We're whittling that down to 10. What's important to me is it will give the city and our team enough information to know what path to put cities on in terms of resources. They can get more customization from us, but for cities, it will be less intimidating.
It also is an intentionality around uplifting stories of cities of all sizes. We are really committed to serving smaller cities, resource-strapped cities, and in these times, our resources are going to get even more tight. It is important for cities to feel like we are accessible. You will also see us intentionally highlighting success stories of smaller cities, so it feels approachable and highlighting stories of cities that when they first came to us, they didn’t have data capacity.
Paterson is one of those cities. They scored a zero, and we could tell that story. They intentionally worked with us, did our capacity building workshops, participated in coaching, organized themselves internally within the city, built their capacity, and then they certified at Silver.
Route Fifty: This whole idea of government efficiency has been in the news for a while and getting a lot of attention in the federal government and states. What does government efficiency mean to you?
HAYNES: When I think about government efficiency, I think about, are governments using their resources in a data-driven and evidence-based way? If government is using data to determine how they're allocating resources, it'll be efficient, because data will show you where you've made investments and where you have not made investments, and so that'll help you direct your resources. Data can also show you where you're getting your desired outcomes from programs and policies that you rolled out and where you’re not. … How are you leveraging that data?
Using data to make decisions is a way of being efficient and that’s part of how I think about efficiency. Efficiency is also about, are you using your resources, including your staff, at the highest and best use? Are staff being deployed to your city's priorities? What are your city's priorities? Have you identified them? And is that where you're focusing? It's being very tactical and practical with your financial resources, but also your personnel resources and leveraging data to do that.
Route Fifty: AI and all these technologies are evolving rapidly. How can cities keep up with that, stay ahead of that, or write policies that reflect how quickly things are changing?
HAYNES: This conversation has changed from last year to this year. Last year, when we were here, our discussion was about what AI is and what cities were using it or piloting it. This year, we all understand there's more understanding of a baseline of what AI is, and now these are ways that cities are using it in very tangible ways, practical ways, whether it's climate, whether it's housing permits, whether it's engaging residents. It's evolving. The technology is rapidly evolving.
One of the things I ask is, within our community, connect with your peers. You do not have to do this alone. You're not meant to do this alone. The field is changing rapidly. Connect with your peers. We highlight the best practices, reach out, learn from one another. If you want to set up some guidelines and governance around AI, you don't have to do it from scratch. There's someone within our community who's already doing it.
You don't have to be the expert, but you do want a baseline understanding and knowledge of it, and figure out, is there a way that AI connects to your priorities? Sometimes when new technology comes out, it's like everyone just has to use it. Sometimes, yeah, use it. You don't want to get left behind, but to what end? Is there a way to connect an AI product to a priority of your city, and will it solve a problem? Know why you want to use it, establish your why, get in community with others to learn best practices and then speak with other experts in the field.
Route Fifty: What's in store for the next 12 months? What are you going to be paying attention to? What's exciting you? What's worrying you?
HAYNES: We're excited. We just hit a milestone of over 100 cities certified in North, Central and South America. This is our 10th year, which is super exciting. I think for us in 2025, we're doubling down on our work in North, Central and South America. We're going to continue to ramp up that engagement. We want to be out there at more events like this one, talking about our work, but also elevating the cities in our network… Front and center this year is about lowering our entry point, keeping the rigor, but lowering the entry point so we are accessible to cities. We're working on doing some analysis that we want to share about how the field has shifted over the last 10 years and what outcomes we're seeing in cities.