Many Detroit residents are still undecided about the government’s potential use of artificial intelligence, with 41 percent of city residents reporting neutral opinion or uncertainty about whether the benefits AI will bring to society outweigh its risks, according to a new survey.
The new issue overview provides an overview of the findings of a representative survey of Detroit residents conducted by the Detroit Metropolitan Community Study at the University of Michigan in Ann Arbor.
More than 2,100 people completed the survey between August 6 and October 1, 2025, and the survey results were weighted to represent the demographics of the city’s entire population.
The survey asked Detroit residents about their general perceptions of AI as well as their support for specific municipal AI applications, such as managing water and wastewater systems, helping residents find government resources, monitoring neighborhood maintenance, identifying criminal suspects, and locating missing children.
“We wanted to know whether residents believe that AI can improve the efficiency, transparency, and accountability of government services, and whether the technology could increase harm, bias, and inequity in service delivery. We hope that city officials will take Detroit residents’ opinions about AI into account as they consider the application of AI to government services,” said UM Postdoctoral Fellow at the Institute for Social Research and AI at the Institute for Social Data and AI. Fellow Hoenuk Ha said:
Mr. Ha co-authored the problem brief with Yucheng Fan, DMACS data manager, and Mara Cecilia Ostfeld, DMACS faculty co-lead.
The survey found that Detroit residents’ support for AI depends on how they intend to use the technology. Support for AI was strongest in identifying missing children, with a majority of residents (57%) expressing support.
Support for AI is much lower in service-oriented applications such as monitoring neighborhood maintenance (38 percent), helping residents access government resources (34 percent), and managing water systems (30 percent).
Using AI to identify criminal suspects received the highest level of opposition (39%). This pattern may partially reflect that 55 percent of residents agree that the use of AI may harm some residents, even if it helps others.
“When we asked Detroiters about AI’s potential to improve government efficiency, transparency, and accountability, our survey found efficiency to be the most compelling benefit of AI, but Detroiters were still widely unconvinced,” Huang said. “Twenty percent to 30 percent of residents were neutral about efficiency gains across different scenarios, suggesting that efficiency alone does not lead to widespread public buy-in.”
Detroiters’ concerns about harm and bias related to AI persist. Across all use cases, a majority of residents (40-55 percent) agree that AI could harm some residents, even if it helps others. Even in the most popular AI use case (identifying missing children), concerns about bias remain high (41%).
Residents are also skeptical that AI will automatically create more equal service provision.
Neutral responses dominate perceptions of fairness, with approximately 30-35% of respondents reporting that they neither agree nor disagree that AI will improve fairness and improve service delivery across use cases. Another 10 to 20 percent of Detroit residents said they were “unsure” whether AI would improve fairness in the delivery of different types of government services.
Support for AI plummets when Detroit residents’ personal data is involved. Almost two-thirds (64%) of residents would not give access to their personal data, and 22% said they would be willing to give access to their personal data in government records to help improve the accuracy of new tools. The remaining 14 percent don’t know where they stand.
“Detroit residents are approaching the use of artificial intelligence in their local government with more caution than enthusiasm,” Ostfeld said. “Many residents remain uncertain about the benefits of these technologies, particularly concerns about data sharing.”
“If the city wants to expand its use of AI, it is important to first build trust by clearly explaining how these systems will be used and how residents’ information will be protected.”
For more information, see the issue summary: Detroiters’ cautious and conditional support for artificial intelligence in local government.
