Massachusetts highlights AI Hub and sandbox to pilot state uses safely

Massachusetts officials told a legislative panel the commonwealth is taking practical steps to deploy artificial intelligence within state government while aiming to protect resident data and civil rights.

Jason Snyder, secretary of technology services and security for the Commonwealth, described a published generative-AI policy, an AI acceptable-use policy for personnel, procurement terms that include AI-specific safeguards, and a private “sandbox” that prevents state data from being incorporated into external models. Snyder said the sandbox is intended to protect sensitive information that would otherwise be exposed by public chat tools.

Snyder also described education and workforce investments tied to the AI Hub. He said the state partnered with higher-education institutions, running student cohorts that collaborated with agencies to develop pilot projects; those projects ranged from engineering-standard lookups for state highway projects to other agency-specific automation that reduced manual preparation time.

Panelists and audience members urged that such pilots be accompanied by training for staff, clear documentation, and oversight mechanisms so that when systems inform high-impact decisions — for benefits, public safety or health care — there are accountable human roles and monitoring.

The official summary of state activity reflected a broader panel consensus: states can shape how federal guidance is operationalized by building procurement standards, workforce training, and testing protocols that raise a baseline for responsible AI deployment while preserving room for innovation.