TLDR: As federal and state governments deliberate on artificial intelligence regulations, numerous U.S. cities and counties are proactively implementing their own AI guidelines for local government workers. These policies, often framed as flexible ‘guidelines,’ aim to enable responsible AI experimentation while addressing concerns like accuracy, privacy, security, and intellectual property. Cities like Boston, Tempe, and Lebanon are at the forefront, with Boston reporting that 60% of its employees use AI weekly for tasks ranging from drafting memos to data analysis, and Tempe seeing up to 100,000 AI uses monthly.
While federal and state governments continue to deliberate on the appropriate use and regulation of artificial intelligence, a growing number of municipalities across the United States are taking the initiative to establish their own AI guidelines for local government employees. This proactive approach is driven by a recognition of AI’s inevitable integration into daily operations and a desire to manage its inherent risks responsibly.
Boston, Massachusetts, stands out as one of the pioneering cities in this movement. In May 2023, Boston’s Chief Innovation Officer, Santiago Garces, and his team rolled out a comprehensive document outlining the purpose, sample use cases, and core principles for generative AI in government work. Garces emphasized the dual nature of AI, stating, “AI is generally useful. But it is a set of technologies that also carries unique risks that need to be considered. And I think that our employees are generally concerned about accuracy, privacy, security and intellectual property.” The city’s strategy was to foster ‘responsible experimentation’ rather than resisting the technology, engaging their workforce as partners in learning.
Boston’s guidelines, which drew inspiration from the Biden-era 2022 White House AI Blueprint, focus on operational use, privacy, protection, and navigating AI’s increasing pervasiveness. The policy explicitly labels generative AI as a ‘tool,’ underscoring that ‘We are responsible for the outcomes of our tools.’ Key principles include empowerment, inclusion, respect, transparency, accountability, innovation, risk management, privacy, security, and public purpose, accompanied by clear ‘dos’ and ‘do nots’ for ethical AI use.
Following Boston’s lead, Tempe, Arizona, released a similar AI policy just a month later in 2023. Stephanie Deitrick, Tempe’s chief data and analytics officer, initiated the framework development after researching data, bias, and machine learning algorithms. The rapid emergence of generative AI tools like ChatGPT prompted the city to implement safeguards. All new AI tools in Tempe undergo review by a governance committee, and relevant state and federal legislation is incorporated. Jared Morris, Tempe’s Director of Information Technology, highlighted the human-centered approach, stating, “These are our values, and we are going to make sure that whatever governance we have aligns with these values.”
Lebanon, New Hampshire, has also adopted a flexible AI policy, designed to adapt to evolving state and city laws. Melanie McDonough, Lebanon’s chief innovation and AI officer, noted, “We’re trying to build a policy that’s robust, that can withstand the pace at which AI is changing. Policy is harder to change. Guidelines can be updated more frequently.”
The adoption of AI tools by local government workers is already widespread. A recent survey by Boston’s team revealed that 60% of employees use AI at least once a week, and 78% expressed interest in learning more about generative AI. Current applications primarily involve drafting memos, proofreading emails, and some data analysis or code generation. Notably, one Boston department utilized Google’s Veo 3 to produce a 20-second video on trash disposal best practices for approximately $30 in credits, a significant saving compared to a preliminary traditional filmmaking quote of around $20,000. Garces remarked on the potential: “You start seeing the potential impact in helping us do things that were either out of our reach or being able to do them faster or being able to do them for less money.”
In Tempe, city employees engage with about 150 different AI applications, reaching a peak of 100,000 uses in a single month. While many are ‘off-the-shelf’ models like ChatGPT for writing and research, others are paid partnerships, such as with AI company Axon for real-time object recognition. This system assists emergency services, for instance, by alerting officials to blue cars if a person in distress is reported in such a vehicle. However, Tempe’s guidelines strictly prohibit facial recognition technology, with Morris emphasizing, “We’re really careful, trying to be very, very careful on anything that could possibly deprive anyone of liberty or job opportunities.”
Also Read:
- New Jersey’s AI Assistant Delivers Significant Time and Cost Savings for State Operations
- Workplace AI Adoption Disparities Fuel Widening Productivity Gaps, Reports Indicate
Cities are choosing to develop their own policies due to the unique accountability they have to their constituents and their ability to act more swiftly than higher levels of government. McDonough stated that operating without an AI policy proved more challenging, and Deitrick stressed that strong governance is crucial to prevent unintended negative consequences. Garces hopes that state and federal regulators will collaborate with cities, leveraging the practical knowledge gained at the local level. “My hope would be that state and federal regulators are working together with cities and not working against them,” he concluded, “Because I think that we have a lot of information and knowledge about how some of these things are starting to occur.”
