csg stack

CSG 50 State Scan Research Series

GREAT EXPECTATIONS: STATE AI POLICY

Section 7: Observations for States Regarding AI Policymaking

AI is a leading policy issue facing states and U.S. territories. Based on the pre-existing applications of AI, the benefits of AI may include increased efficiency in government operations along with reduced tax and administrative burdens. In addition, this report has examined AI legislation at the federal, state, and territory level in a structured and comprehensive manner.

Furthermore, this report introduced the CSG State AI Competitiveness Indicators, a dataset that compiles the latest information on AI-related measures and shows where states and territories are succeeding in AI innovation, development, and readiness.

These indicators, along with case studies of successful AI policymaking, are intended to inform state leaders and provide options for decision making for a variety of economic contexts.

The analysis of these indicators reveals existing space for regional cooperation between states on AI policy, such as research collaboration and integrating AI infrastructure. Also, these indicators provide insight into critical factors that businesses consider when deciding to build or relocate AI firms.

Useful Principles for States and Territories Regarding AI Policy

Looking to the future, several principles and practices emerge from the present analysis of the AI sector. Where appropriate, states can learn from each other to facilitate prudent decision-making on AI issues. A few considerations for state leaders include:

  1. State lawmakers and their constituents should decide the trajectory of AI and emerging technologies within their state.

  2. Key stakeholders ought to be mindful that AI initiatives that flourish in one state might be unpopular in another setting. For example, data centers hubs have taken root in Northern Virginia, while other states are presently debating the costs and benefits of these IT infrastructure projects.

  3. Foster public-private partnerships to examine how AI solutions might be utilized in government operations. State governments can scrutinize use cases (chatbots on government websites, pilot programs of Gen AI solutions on government workforces) to determine what works best for their specific circumstances.

  4. Establish clear rules that govern the ethical use of AI and enforce rules that strongly discourage the harmful usage of AI. This might include the responsible collection of data used to train AI models, discouraging individuals from using AI for deepfakes, or governing the spread of false information generated through AI-related means.

  5. Think about human-centered AI systems and policies. States and territories are asked to factor in the impact of AI on local communities, whether it be in terms of long-term changes to employment outcomes or broader changes in the regional economy.

State Legislator Takeaways

In addition, our recent convenings with state legislators across the country emphasized the following major takeaways:

1. Transparency and Accountability Are Cornerstones of AI Governance

Legislators stressed the public’s right to know when and how AI is used in consequential decisions.

Action Implications:

  • Require disclosure when AI tools influence hiring, healthcare, education, or public benefits decisions.
  • Mandate audits and third-party testing to identify bias or malfunction.
  • Introduce adverse event reporting systems for AI failures.
 
2. Human Oversight Must Remain Central

Consensus emerged that AI should augment, not replace human judgment.

Policy Actions:

  • Require human sign-off on critical healthcare, education, and government decisions.
  • Prohibit mandatory AI use in schools or public services without human review.
  • Ensure accountability for outcomes remains with people, not algorithms.
 
3. Consumer and Citizen Protection Is Foundational

AI systems can perpetuate bias and discrimination without strong guardrails.

Policy Actions:

  • Define and prohibit algorithmic discrimination in law.
  • Create enforcement mechanisms for transparency and fairness.
  • Protect children and vulnerable groups from harmful AI applications and misinformation.
 
4. Workforce Impacts Demand Immediate Attention

AI disproportionately affects younger and entry-level workers, particularly in technology and service sectors.

Action Implications:

  • Invest in retraining and workforce transition programs.
  • Partner with employers and educational institutions to identify emerging skill needs.
  • Track AI’s labor market impacts to inform targeted interventions.
 
5. Environmental Sustainability Cannot Be Ignored

AI’s growth is driving energy and water consumption through data centers.

Policy Actions:

  • Standardize reporting on data centers’ water and energy use.
  • Coordinate between energy and water agencies to address local impacts.
  • Promote use of non-potable water and advanced cooling technologies.
 
6. States Are Leading the Way

With limited federal regulation, states are shaping the AI governance landscape.

Next Step: Expand interstate collaboration to share model legislation and implementation lessons.