Maryland Gov. Wes Moore released a statewide artificial-intelligence agenda Tuesday that pairs proposed legislation with agency directives on civil rights, worker protections and child safety. The framework is organized around three principles: protecting residents, giving workers and unions a voice, and shielding children from exploitation and harm.
The plan is a policy roadmap rather than one new law. Some steps can begin inside executive agencies, while others will require cooperation from the General Assembly. Among the legislative proposals, Moore directed the state’s AI Subcabinet to develop recommendations for regulating frontier-model companies. The administration says those recommendations should address safety testing, independent audits, whistleblower protections, public reporting and 72-hour incident notices.
Other proposed measures would create a statutory property right in a person’s likeness, restrict landlords from using shared nonpublic competitor data to coordinate rents, and limit gambling operators’ use of algorithms to target people showing signs of problem gambling. State civil-rights, labor, housing and financial regulators would also issue guidance explaining how existing protections apply when automated systems influence employment, lending, housing or other consequential decisions.
For state employees, the administration plans a standing labor-management committee to examine how AI affects work. The broader agenda also calls for training and transition support, human oversight in consequential personnel decisions and restrictions on social scoring, covert biometric identification and tracking intended to undermine organizing. The state says its goal is to use AI to augment public workers rather than replace their judgment.
The children’s section calls for stronger rules for chatbots, including crisis protocols for suicide-related or other distress signals and added friction for users under 16. Moore also wants human-escalation procedures for AI tools used in schools, disclosure to families, evidence that classroom products improve learning and model lesson plans on safe and ethical AI use.
Maryland already has a government-specific foundation for these proposals. A 2024 executive order established principles of fairness, innovation, privacy, safety, transparency and accountability for agency AI use. The Department of Information Technology’s Responsible AI Policy subsequently required impact assessments, bias testing, human oversight and periodic review for high-risk state systems, while prohibiting uses such as social scoring and unlawful surveillance.
The new agenda expands beyond state procurement and operations toward private-sector conduct and schools. That shift is consequential, but it is not self-executing: details such as which companies qualify as frontier developers, what audits must show and how violations would be enforced remain to be written. The next General Assembly session will determine which proposals become binding statewide rules and which remain administrative guidance.
Maryland’s approach also continues work begun when Moore created the AI Subcabinet. At the time, Associated Press reporting described the initiative as a first step toward guardrails for agency systems. Tuesday’s framework provides a broader list of targets; the next measurable test will be whether agencies issue the promised guidance and lawmakers translate the proposals into enforceable text.
