Video not working? Click here to view the webinar on YouTube.
Since 2021, cities have tapped innovations in artificial intelligence and other advanced technologies to plan, deliver, operate, and maintain high-impact infrastructure projects. These procurements present many new challenges for cities. And now, both the technology and policy landscape are shifting rapidly. This creates a complex challenge: how can cities balance innovation with prudent technology risk management?
This roundtable session brought together experts in federal, state, and city technology procurement to share fresh insights and tested strategies for moving technology innovation projects forward under uncertainty and volatility. As federal infrastructure and AI policy both go through major overhauls, how can cities anticipate and adapt to continued disruption? When does it make sense to move forward with AI procurements in infrastructure projects, and what measures can cities take to reduce risk? When should cities seek options to wait and see, and what are the costs of deferred delivery for AI innovations?
Insights from the discussion included
Waiting out the disruption is a wise strategy.
Waldo Jaquith, Government Delivery Manager, U.S. Digital Response, noted that the market for AI is still unsettled—and cities that move too quickly to procure solutions could be left holding the bag. He cited the case of Chinese AI startup’s Deepseek R1 mode which, when released just a few months ago, triggered a new round of cost-cutting measures by market leaders OpenAI, Google, and Anthropic. Jaquith also noted that ongoing efforts to replace federal workers with AI—and the inevitable failures that will follow such hasty and widespread experimentation—could taint government AI projects for years to come.
Access AI through software you already own.
A middle road strategy is to take up the opportunity to leverage AI features being rolled out in existing software products and service cities already own. Santiago Garces, Chief Information Officer, City of Boston, Massachusetts pointed that as cities’ current vendors will integrate AI into their existing
offerings, procuring AI in government might look like procuring “cut and paste”. Even cities looking for deeper capabilities can procure them from cloud providers without an RFP because they are already listed on a NASPO or a GSA schedule.
Look out for hidden liabilities in everyway AI use.
Today’s AI use is generating new kinds of technical debt. Katy Ruckle, Chief Privacy Officer, Washington State highlighted the importance of weighing cost versus benefits, especially in high-risk AI deployments such as risk scoring decisions or facial recognition. Even mundane, everyday AI uses contain hidden risks. Prompts sent to AI chatbots may be subject to transparency laws. And using AI to transcribe meetings can capture non-public information at the beginning and end of recordings, creating new workloads for public workers that need to go back and make redactions.
About the Tech and Innovation Center Series (T&IC)
The T&IC series is dedicated to helping local leaders navigate and understand the large quantities of information from the federal government on the nearly 400 funding opportunities available through the Bipartisan Infrastructure Law. The 2 year series is focused on how cities can leverage technology to improve their federal infrastructure funding proposals during 2023 and 2024. Programs focus on helping cities improve their proposals in response to Notices of Funding Opportunities (NOFO’s) by adopting state-of-art technologies, expanding their technology capacity, and integrating aspirational technology “moonshots” for their cities.
The Series is produced by the Jacobs Urban Tech Hub at Cornell Tech and U.S. Digital Response (USDR) as part of the Local Infrastructure Hub, a partnership of the US Conference of Mayors, National League of Cities, Results for America and Delivery Associates supported by Bloomberg Philanthropies, Ballmer Group, Emerson Collective, Ford Foundation, and the Kresge Foundation.