California, New York, New Jersey and North Carolina are rethinking artificial intelligence oversight, prioritizing public benefit and risk over industry competition as new systems enter workplaces and public services.
As artificial intelligence continues to advance, several states are moving away from the idea of an AI "race" and instead focusing on how these technologies affect residents, workplaces, and public services. Rather than asking which company or country will win, state leaders are examining who benefits from AI, who bears its costs, and who is at risk when systems fail.
Major AI companies have recently adopted the term "pacing" to describe slowing development when safety measures lag behind technological progress. Some have paused work on new models due to cybersecurity concerns, released lengthy reports on their safeguards, and called for government involvement to ensure AI does not outpace society's ability to manage it safely. While these steps are important, state officials argue that responsible AI oversight must go further, addressing the real-world impact on people today.
States Lead With People-Centered AI Policies
California, under Gov. Gavin Newsom, is studying AI's effects on employment, monitoring early signs of workforce disruption, and analyzing whether certain demographic groups are disproportionately affected. In New York, Gov. Kathy Hochul has established an independent advisory board that brings together leaders from technology, labor, nonprofits, finance, health care, and civil society to guide responsible AI adoption and assess its impact on the workforce and public-interest applications.
New Jersey is taking a hands-on approach with its NJ AI Assistant, an internal chatbot now used by about 20,000 state employees across more than 300,000 sessions and one million prompts. North Carolina has formed an AI Task Force to evaluate how state agencies use AI and to help shape a statewide policy that puts people first.
Measuring Real-World Impact
These state initiatives share a common principle: AI oversight should begin and end with people. This means involving public employees before introducing new tools, testing systems with those who rely on government services-especially groups historically excluded from design processes-and measuring whether new technology reduces wait times, errors, abandoned applications, and unnecessary paperwork.
For most Americans, AI is not an abstract competition. It is a tool that may influence whether they receive a job, a loan, or a public benefit, or it may be the reason their electricity bill changes. As AI becomes more integrated into daily life, states are working to ensure that those affected have a voice in how these systems are designed, developed, and deployed.
Looking Beyond the AI Race
While federal policymakers and industry leaders debate the future of AI, state governments are focusing on immediate, practical impacts. Their efforts reflect a shift from asking who will win the AI race to asking who benefits-and who might be left behind. This approach is echoed in other areas of technology policy, such as election security, where state officials have sometimes clashed with federal agencies over priorities and resources, as seen in reporting on state election security concerns.
By prioritizing public benefit and risk, states are setting a new standard for AI oversight-one that centers on the needs and experiences of their residents rather than the ambitions of technology companies.
Amanda Renteria is CEO of Code for America.
