At WHOOP’s Boston headquarters last night, Governor Maura Healey unveiled a plan to put an AI assistant in the hands of roughly 40,000 Massachusetts executive-branch workers, casting the move as a way to speed up the drudgery of government work so humans can focus on the tougher stuff.
The tool is powered by ChatGPT and will roll out agency by agency, with training and guardrails promised along the way so staff learn what the assistant can, and very much cannot, do. Healey framed the moment as a turning point for state government.
“We’re at a moment where things can get much better quickly,” she told the crowd, while OpenAI CEO Sam Altman joined remotely to walk through how the tool could help with communications, document drafting and new automated workflows.
Under a three-year contract, the state will pay OpenAI a per-user fee that starts at $13 per month and falls toward $9 as more workers come on board. State officials say the app will operate in a walled-off environment so that employee inputs are not fed back into OpenAI’s public models. The Executive Office of Technology Services and Security will be the first to test-drive the assistant starting Friday, with other agencies phased in over the next several months, according to The Boston Globe.
State Investment And The AI Hub
The AI assistant is one piece of a larger push to brand Massachusetts as a national home base for applied artificial intelligence. In 2024, Healey created an AI strategic task force and later secured about $100 million from the Legislature to stand up an AI hub aimed at boosting research, workforce development and safe testing grounds.
The hub is meant to knit together universities, startups and public agencies, while offering secure “sandbox” environments where new tools can be tried without putting sensitive state data at risk, according to Mass.gov.
Private Sector Partners
Healey’s announcement came during the launch of the Massachusetts AI Coalition, a new private-sector effort that wearable tech company WHOOP helped organize. The coalition aims to pull together local companies, investors and builders who are eager to experiment with AI in real-world settings.
Organizers say they plan to host dozens of workshops, hack days and networking events designed to speed up AI adoption across the Commonwealth, according to WHOOP.
Guardrails And Politics
The timing is not accidental. Beacon Hill is in the middle of debating how to rein in some uses of AI in politics, including proposals that would outlaw misleading AI-generated media in campaigns and require clear disclosure when synthetic content is used.
State officials are also reminding workers that chatbots can produce flat-out wrong or “hallucinated” information. That tension, racing to harness AI’s potential while slapping on enough guardrails, is shaping how the administration plans to train employees and monitor how the new assistant is used, according to The Boston Globe.
What To Watch Next
Massachusetts is hardly the first government to test an AI sidekick. Other states that have piloted similar tools report that some agencies are saving time, while others are running into accuracy problems and cost concerns.
Experiments in California, Pennsylvania and Colorado have shown productivity gains in certain offices, but the mixed results highlight the importance of strong data governance, human review and serious training for workers, according to reporting by The Pew Charitable Trusts. Healey’s team says training will be built into the rollout and that the state will track outcomes as more agencies come online.
