Brief Artificial Intelligence and Health Care: A Primer

Artificial Intelligence and Health Care: A Primer 

Policymakers are increasingly turning their attention to the use of AI in the health care sector. This emerging technology brings promising applications—as well as considerations around potential risks—especially in fields as consequential as health care. State legislators are considering applying various guardrails to the use of AI in health care related to privacy, data integrity, transparency and more to ensure AI’s safe and effective use. 

NCSL’s Health Innovations Task Force explores issues and opportunities faced by state policymakers as they reform their own health system. In 2024, the task force explored the emergence of AI in the health care sector. This resource summarizes concepts discussed at recent meetings, including a glossary of terms, examples of applications of AI in health and state legislative themes.  

Glossary of Terms

The term artificial intelligence, sometimes referred to as automated decision-making, serves as a broad umbrella label that includes various subsets of AI, e.g., machine learning, robotics, etc. However, despite the emergence of AI as a promising new tool, there is still no consensus on a singular definition. The lack of an overarching definition is challenging to lawmakers as they seek to create a regulatory framework. Nevertheless, policymakers need to settle on some guiding principles to continue progress. The definitions below are just one set among many.

Artificial Intelligence: A machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments. Artificial intelligence systems use machine and human-based inputs to (a) perceive real and virtual environments, (b) abstract such perceptions into models through analysis in an automated manner; and (c) use model inference to formulate options for information or action.

Machine Learning: A subset of artificial intelligence that automatically enables a machine or system to learn and improve from experience. Instead of explicit programming, machine learning uses algorithms to analyze large amounts of data, learn from insights and then make informed decisions.

Deep Learning: A subset of machine learning that uses artificial neural networks to process and analyze information. Deep learning algorithms are inspired by the neural networks of the human brain and are used for analysis of data with a logical structure.

Predictive AI: Artificial intelligence systems that utilize statistical analysis and machine learning algorithms to make predictions about potential future outcomes, causation, risk exposure, and more.

The large majority of AI used in the health care sector falls into the category of machine learning, but some other types of AI include narrow AI, natural language processing, computer vision and robotics.

Applications of Artificial Intelligence in the Health Care Sector

Scientists have been exploring using AI in medicine as early as the 1970s. This research resulted in various AI applications being used in present-day health care. These applications may include automating administrative tasks, assisting in clinical care, informing diagnoses, improving precision medicine, predicting patient risk for certain adverse health events, assisting with drug discovery and much more.

State Legislative Themes

States are increasingly introducing legislation related to AI and health. In 2023, 11 states introduced legislation specifically related to health AI-an uptick from three states in 2022. This trend has continued into the 2024 session.

States are considering a wide variety of policy options in this area. Some of the legislative themes from these past legislative sessions include:

• Studies: Legislation requiring a study of artificial intelligence or creating a task force, advisory body, commission or other regulatory, advisory or oversight entity. 
• Responsible Use: Legislation that may prohibit use of AI tools that contribute to any type of algorithmic discrimination, unjustified differential treatment, or impacts disfavoring people based on their race, color, ethnicity, sex, religion, age, national origin, language, disability, veteran status, genetic information, reproductive health, or other classifications protected by state laws. 
• Patient Notification: Legislation that requires patients to be made aware of the use of AI and, in some cases, allow patients to opt-out of its use.
• Practitioner Monitoring: Legislation requiring providers or administrators to monitor AI tools that are being used. In some cases, providers or administrators may be tasked with decision validation and/or possess the power to override AI when deemed appropriate.
• Developing Standards for Developers and/or Deployers: Legislation creating standards for developers and/or deployers of AI. This type of legislation tends to be more general in scope, but many states have distinguished high-risk AI systems when crafting their regulation which often includes AI systems being utilized in health care.

Additional Resources

• Artificial Intelligence 2024 Legislation (NCSL)
• Approaches to Regulating Artificial Intelligence: A Primer (NCSL)
• Assurance Standards Guide: Executive Summary (Coalition for Health AI)