Leaders have ‘urgent responsibility’ on AI regulation and use, report says

States’ efforts to regulate artificial intelligence were dealt an apparent blow at the end of 2025, as President Donald Trump signed an executive order seeking to preempt them.

In an order he said would target “cumbersome” state AI regulation, Trump said states deemed to be stifling innovation could face restricted grant funding. It’s an effort that is likely to face legal challenges.

Meanwhile, U.S. Sen. Ed Markey, a Massachusetts Democrat, introduced legislation at the end of last year to block Trump’s order as part of the ongoing appropriations process. In a statement, Markey called Trump’s executive action “a direct threat to our children, to workers, to our planet, and to marginalized communities.”

And as states look to keep regulating AI in the face of federal threats, a new report said policymakers at the state level have a critical role to play in ensuring that various safeguards around AI are in place to protect residents’ privacy and security. Those same lawmakers have a tremendous opportunity to take advantage of the technology to make government function better and deliver better results. The report was produced by the nonprofit NewDEAL Forum’s AI Task Force, which convened elected leaders, nonprofits and the private sector to determine a path forward.

“Across our conversations, one lesson has been clear: governments at every level cannot afford to sit on the sidelines as passive observers or focus solely on regulation,” says a message in the report from Albany, New York Chief City Auditor Dorcey Applyrs, New York Assemblymember Alex Bores and San Jose, California Mayor Matt Mahan, the task force’s co-chairs. “[Our] charge as elected officials is to ensure that AI works for the people we serve — transparently, equitably and responsibly.”

The report says AI can help government work smarter, as it can help leaders anticipate their residents’ needs and deliver support before issues escalate. AI can help address longstanding challenges like outdated workflows, slow response times and fragmented data, the report says, adding that it can make government services “more responsive, equitable, and effective for the people they are meant to serve.”

But to get there, NewDEAL said agencies must partner strategically with the likes of innovation offices, technology providers and other civic organizations so that systems integrate services effectively. The group also said systems must be designed for users first to reduce complexity, and be piloted on a small scale before they are expanded.

Some states and localities have been successful, too. NewDEAL pointed to the likes of New Hampshire and Maryland, which have found success in using AI to streamline the public benefits process and get help out the door quicker. Governments across the country have also used the technology to streamline permitting, make their communities safer, optimize traffic safety and infrastructure and do better planning and community engagement.

Other policy areas, like ensuring inclusive healthcare access and improving education outcomes, are still in the very early stages of determining how AI can be helpful, but NewDEAL said there is optimism there too. And while many government workers are ready for AI, they need support and training, something that states are slowly embracing with various upskilling initiatives and are looking to accelerate this year.

“Public servants are stretched thin, often facing rising demand without additional resources,” the report says. “Artificial intelligence offers an opportunity to reduce workloads, improve training, and preserve institutional knowledge — but only if workers are part of the design and deployment process. Governments that invest in their employees are already seeing early returns — from faster service delivery to stronger morale and innovation.”

Despite the optimism, there are plenty of pitfalls that governments must navigate around AI. The report says the public must be protected from malicious actors who may use the technology for deepfakes, while strong consumer protections will ensure humans make decisions on critical services, not machines. And NewDEAL noted that several states are “advancing thoughtful approaches” to regulating so-called “frontier models” of AI, which represent the most powerful systems that are evolving the fastest.

More lies ahead, however, including protecting democracy and election integrity, given that AI could “amplify misinformation and impersonation,” the report says. The technology also faces several adoption barriers among the public sector, NewDEAL said, including a lack of resources and data literacy, a culture of resistance among some staff, and worries about privacy and digital equity.

In sum, NewDEAL said that while AI has shown plenty of promise in helping state and local governments move towards a brighter future, a great deal of work lies ahead if they are to help the technology realize its full potential while keeping residents safe.

“Amidst unpredictable federal action around AI regulation, state and local leaders have an urgent responsibility to model people-first AI governance and deployment,” the report says. “The future of government is being designed now, and it is our obligation to ensure that AI serves the public good.”