Published by the City of Boston · December 2025

In a 2025 City of Boston employee survey, 78% wanted formal generative-AI training, and 60% had already tried the technology.

Published at boston.gov

In spring 2025, the Boston Analytics Team and Office of Emerging Technology surveyed more than 20% of the City's general workforce on generative-AI use. 60% had used the tools at least once and 25% were using them weekly at work. 78% wanted formal training, with about half describing themselves as very interested; 42% reported lower confidence with new digital tools generally.

City of Boston · employee generative-AI survey · 2025

Demand for generative-AI training outran adoption among City of Boston employees.

Spring 2025 survey of more than 20% of the City of Boston's general workforce, excluding Boston Public Schools, Boston Fire, Boston Police, Boston Public Health Commission, and Boston Public Library. Mixed multiple-choice and open-ended items on use frequency, training preferences, and concerns about generative-AI adoption.

78%

of surveyed City of Boston employees wanted formal generative-AI training.

About half described themselves as "very interested."

60% HAD TRIED GENERATIVE AI · 25% USING IT WEEKLY at work.

SAMPLE: more than 20% of the City's general workforce, spring 2025.

Source: City of Boston Analytics Team and Office of Emerging Technology, spring 2025 employee survey · sample: more than 20% of City of Boston employees, excluding Boston Public Schools, Boston Fire, Boston Police, Boston Public Health Commission, and Boston Public Library · aggregated results and questionnaire published December 2025 alongside the 2023 Generative AI Guidelines.

Methodology

The survey was administered by the Boston Analytics Team and the Office of Emerging Technology within the Department of Innovation and Technology, with support from Boston Digital Service. More than 20% of the City's general workforce responded. Uniformed services (Boston Fire and Boston Police), Boston Public Schools, the Boston Public Health Commission, and the Boston Public Library were excluded from this iteration. The instrument combined multiple-choice and open-ended items on use frequency, training preferences, intended use cases, and concerns about generative-AI adoption.

Use today

60% of respondents had tried generative AI at least once. 25% reported using it weekly at work. As a baseline for adoption, 42% of respondents reported lower confidence with new digital tools generally, and 25% reported lower confidence with search engines, indicating that a meaningful share of the workforce begins from a non-power-user starting point.

68.5%

Inaccuracy was the top workforce concern about generative AI

57.6%

Security ranked second

50.6%

Data privacy ranked fourth, behind plagiarism/IP at 53.8%

Concerns

Workforce concerns concentrated heavily on inaccuracy and security. Beyond the three in the strip above, 53.8% flagged plagiarism or intellectual-property concerns, 35.8% bias, 33.1% job displacement, and 30.6% ill-intent or unethical use. Roughly 8% of open-ended responses raised the environmental impact of generative AI, a concern the survey did not explicitly prompt for.

Training demand

78% of respondents said they were interested in formal generative-AI training, with about half (~50%) describing themselves as very interested. The findings inform the City's training curriculum and use-case prioritization, alongside the 2023 Generative AI Guidelines that remain in effect.

Read the full piece at boston.gov Includes the full questionnaire, the aggregated results spreadsheet, and additional context on the City's use of generative AI to support Boston residents.