Don’t CON Law Artificial Intelligence

Don’t CON Law Artificial Intelligence

Artificial Intelligence (AI) is quickly entering the health care system, and policymakers are trying to figure out how to respond. AI has huge potential to improve access, costs, and quality of health care, but rising concerns have already led some lawmakers to consider sweeping regulations before any benefits can be realized. We risk of repeating policy mistakes that have contributed to many of the issues we face in health care today.

A recent article highlighted how AI is expected to increase health care costs next year because it will more thoroughly document and charge for services and likely allow more products and services to flood into the market, increasing spending. This is very similar to the reasoning used to justify Certificate of Need (CON) Laws 60 years ago.

CON laws are a long, burdensome process where medical facilities must prove to the state government that their services are necessary to be able to open or expand. It can cost tens of thousands of dollars to apply, take years to complete, and there is no guarantee of a favorable outcome. The intent was to make health care more affordable and accessible through government central planning. It should come as no shock that CON laws had the opposite effect by limiting supply, increasing costs, and worsening health outcomes.

When state governments began implementing CON laws in the 60’s, why did they think that limiting health care facilities would make health care more accessible? The answer is the rising use of technology. The common belief was that having too many hospital beds or extra technological equipment would lead to unnecessary use and higher spending.

This almost makes sense when you think about all the technological advancements taking place at the time. Computers brought about electronic medical records and information systems. Modern imaging, like ultrasounds and CT scans, began being used in clinical settings. Therapeutic devices like pacemakers and defibrillators were more reliable and widely used. One could imagine how this technological boom would have come with many similar concerns we have now with AI, like higher costs, data security, and patient safety. But we don’t have to repeat history.

There is so much potential for AI to reduce health care costs, lowers prices, and improves accessibility and outcomes. For example, Dartmouth has developed a therapy chatbot where 90% of the responses are consistent with therapeutic best practices, and it reduced symptoms of depression by 51% for participants in the first clinical trial. This could be an excellent 24/7 resource that can reduce burden on providers and be particularly helpful in rural areas, for low-income earners, and those who otherwise fall through the cracks of our traditional mental health care system. That is, if we let it.

Of course, concerns with AI should be seriously examined, but lawmakers must resist falling into the traps of broad bans and burdensome requirements. Vermont has already passed a law that almost completely bans the use of AI in mental health services, only allowing it to streamline administrative tasks. So, its residents will likely never benefit from tools like the Dartmouth therapy chatbot, and the only people allowed to use AI are incentivized to increase the use of, and thoroughly bill for, their services—just like that article warns.

We can protect consumers without artificially limiting positive outcomes, but regulations must be targeted and evidence-based. ALEC’s new model policy, the Artificial Intelligence Mental Health Transparency and Accountability Act, tries to do just that. It requires transparency and clarifies liability so that consumers are protected, and bad actors are held accountable. But it also establishes a voluntary safe harbor so that well-intentioned innovation isn’t punished, and tools can continually improve. Innovation and safety are not opposite goals, and this policy is an example of how both can be thoughtfully supported.

The question isn’t whether AI will transform health care. It’s whether policymakers will allow that transformation to benefit the people who need it most. We cannot make the same mistake today that CON laws did 50 years ago. If lawmakers respond to early concerns by restricting innovation before it has a chance to prove its value, they will lock in the same structural problems that already make care unaffordable and inaccessible. Rather, states can and should encourage responsible development to ensure that AI becomes a tool that expands access, reduces costs, and strengthens the health care system.