NJSPL Report Release: AI Policies and Scientific Expertise in U.S. State Legislatures

Science and technology are key aspects of many of the social and cultural challenges state governments face on a daily basis. The list of science-adjacent issues facing states is almost innumerable: climate, health care, media technology, and social media, to name a few. But perhaps no issue challenges state governments more than the rise of artificial intelligence (AI). While the rapid increase in AI capability has promising aspects, there are also significant dangers. States must be well-positioned to both use and defend against AI along two trajectories. States need legislators and government officials with sufficient technical knowledge to understand the potential benefits and dangers of AI, and a sufficient AI infrastructure to provide both opportunities to use the technology and guardrails against bias, poor information, and unforeseen consequences.

The AI Policies and Science Expertise in U.S. State Legislatures project, starting in October 2024 and ending in the summer of 2025, explored connections between both trajectories. More specifically, the project researched STEM expertise among state-level policymakers and legislative activity as well as state preparedness in the artificial intelligence field. The project accomplished this by taking a two-pronged approach to consider both needs: first (1) we expanded and updated an existing database of all STEM and healthcare experts who are elected members of U.S. state legislatures; and second (2) we conducted a landscape analysis of AI-related preparedness at the state level to create a framework to measure how prepared each U.S. state is to adopt, develop, and leverage artificial intelligence technologies efficiently.

Note that the project originally intended to establish a database of all state legislation on AI. However, the National Conference of State Legislatures (NCSL) launched a similar effort soon after the beginning of this project, leading us to pivot to consider a second-level question: how to determine and display state AI preparedness.

The work along both trajectories led to a series of policy recommendations:

• Based on our research, state legislatures broadly lack a significant number of lawmakers with science, technology, and health care backgrounds. This is particularly true in technology fields. Very few state legislators have background in AI technology or the groundwork to understand how to legislate the issues. It is therefore crucially important that: (1) states supplement AI and other technological knowledge in other offices, including legislative staff, so that the expertise is available to legislators and executive policymakers, and (2) states support public efforts to encourage scientists to enter civic spaces and pursue careers in civic science.
• Efforts are undertaken to increase the number of scientists running for public office, particularly at the state legislature level.
• States must create strong infrastructure around AI issues to ensure that each state is prepared for the opportunities and challenges AI creates. We determined, based on the creation of a readiness index as a result of our research, that while some states are much more prepared to handle AI issues than others, on the whole, states need to increase their capabilities across four pillars (more information about each pillar is available in the full report): Government, Data, AI Workforce, and Infrastructure.
• Visionary leadership, comprehensive privacy laws, rich research ecosystems, and public sector capacity are among the attributes generally shared by the states most prepared for AI issues.
• Cybersecurity infrastructure and broadband access are among the current issues most likely to create bottlenecks in AI preparedness.
• More data governance frameworks are needed to avoid risks to transparency, equity, and AI trustworthiness.

Authors