Maryland AI survey puts worker training at the center of adoption
A new state survey separates widespread AI use from deeper workplace integration.

Maryland released its Business AI Benchmark on October 1, bringing new employer survey results into the state’s debate over worker training and AI governance. The survey reached nearly 300 senior decision makers during June and July.
Among respondents, 91% reported some AI use, while 58% remained at a basic level of adoption. Among regular users, 92% reported a positive productivity effect, although most described that improvement as slight.
Expectations are not employment outcomes
The state says 64% planned to have existing employees do more with AI rather than add staff. Only 4% expected a smaller workforce. Those figures describe respondents’ expectations, rather than measured job losses or gains across Maryland.
The policy discussion predates the survey release
The findings arrive after Maryland’s September 22 AI framework. That agenda called for worker participation in adoption decisions, training and transition support, and recommendations for frontier AI oversight. It also proposed bringing unions, businesses and higher education together to examine changing jobs and skills.
The framework describes an agenda for further work. It should not be read as proof that every proposed protection has already become an enforceable rule. The October survey release provides a new input into those policy choices.
The workplace question is what changes next
A useful response to rising AI use is to ask employees which tasks change, what training they need and how the benefits will be assessed. Adoption rates alone cannot answer whether jobs become better or whether workers gain a meaningful say in the transition.
Respondents also asked for AI training, funding for smaller firms and affordable access to secure digital infrastructure, according to the state’s summary.
Related coverage explores the changing workload around AI document review.
Featured image is an original AI generated conceptual editorial illustration.



