Fifty-nine state agencies are now using Google Workspace with Gemini, Google’s generative AI platform, as of Wednesday. The integration via Google Public Sector provides nearly 43,000 state employees with access to Gemini’s AI tools, and 12,500 active users are already testing its capabilities in their daily tasks.
Officials said the central goals are straightforward: boost productivity, reduce manual work, and encourage creative problem-solving — without sacrificing security or ethical standards.
Through the new tools, agencies are experimenting with everything from building no-code chatbots and drafting reports to analyzing sentiment in data and creating website mockups on demand. For employees, this means AI can handle the first pass on many repetitive tasks — allowing them to focus on higher-value projects and constituent services. Maryland has an existing Google Workspace subscription and the rollout of Google Gemini came at “no additional cost” to the state.
According to Nishant Shah, senior adviser for responsible AI at Maryland’s Department of Information Technology (DoIT), the state’s partnership with Google directly supports the Gov. Wes Moore administration’s strategy to responsibly scale AI use across government, as “Maryland is on a journey to adopt AI responsibly, ethically and productively to improve constituent outcomes.”
Maryland’s 2025 AI strategy is, Shah said via email, focused on building momentum around “experimentation and adoption,” recognizing that the technology landscape is still evolving rapidly — from tool capabilities and business models to emerging risks and opportunities. Getting generative AI tools into employees’ hands, supported by “strong security, privacy, and risk controls” and paired with foundational AI training, is crucial, Shah said, to ensure the state workforce can “learn by doing.”
He noted that implementing new tools always carries certain costs — from training to governance and compliance. To manage that, DoIT recently established the Office of AI Enablement, which helps draft policies, facilitate pilot initiatives and foster AI adoption across state agencies.
“They work collaboratively across our agency and with our agency partners to make tools like Gemini available to the state workforce, working directly with them to identify and execute real workflow improvements,” Shah said. “We provide free AI training tailored for civic servants through our partnership with InnovateUS, which we secured at no cost beyond staff time to operationalize the resource.”
The state, he said, will be evaluating and offering more role-specific AI training in coming months and years.
As AI use grows, early returns are already emerging. Shah said about 30 percent of eligible employees use the tools regularly.
“There are specific instances where experimentation is already leading to significant savings,” he said. “For example, our partners at the Maryland Department of the Environment have already saved more than $400,000 by utilizing Gemini Google Workspace to fulfill internal tasks for their water infrastructure grant application rather than procuring a vendor solution.”
He added that AI applications beyond Gemini are also delivering measurable results. One example is the Maryland SUN Bucks program, which provides summer grocery assistance to families of school-aged children receiving free or reduced-price meals.
“To learn more about the program, Marylanders can access a custom-built AI chatbot, called ‘SUN Bucks Chat,’” he said. “It’s fully functional in English and Spanish, can be accessed from a laptop or smartphone, and allows users to restart chats as needed.”
Over 16,000 users have used the tool, according to Shah, and research by the Maryland state benefits team found it accurately answered questions in 97 percent of scenarios — “saving significant staff time.”
All AI projects in Maryland, he noted, are classified by risk tier under the Responsible AI Policy — to make sure proper oversight exists.
The state’s road map for AI remains focused on five key areas: governance, data foundations, experimentation, literacy and domain-specific strategy, Shah said.
“For 2026, we plan to help the state move up our maturity curve — from pure foundation building in 2024 to momentum building in 2025 to a focus on enterprise adoption and use in 2026,” Shah said. “That includes more clarity on the pathways agencies have to go from idea to experimentation to adoption, expanding AI literacy offerings, and embedding AI expertise more deeply across agencies.”
