Skip to content

The AI

Presented by

Health Care & Scientific Development

Health Care & Scientific Development

The health care sector is on the verge of big changes as artificial intelligence reshapes how researchers, drugmakers and providers operate. 

The possibilities are endless, from speedier drug development to lighter paperwork burdens for providers. But policymakers are also concerned about the potential pitfalls in an industry that deals with human lives and sensitive personal data.

As the health care industry embraces AI, we examine how policymakers in Congress and the federal government are approaching the new reality, their concerns and what they envision for a regulatory framework.

Be sure to listen to our accompanying podcast as we dissect the issue and check out the rest of our AI Impact series here.

Enabling Innovation in Dermatology


Health datasets play a crucial role in research and medical education. To address this, Google collaborated with Stanford Medicine physicians to release the Skin Condition Image Network (SCIN), an open-source dataset of skin condition images. Today, scientists and doctors can use SCIN to advance dermatology research, develop better skin condition diagnosis tools, and improve medical training.


Learn more

Current State of Play in Washington

As with any new technology, Congress and federal agencies have to craft a regulatory regime that balances AI innovation with standard-setting.

Congressional committees with jurisdiction over health issues have held hearings in recent months to debate the risks and rewards of AI as they seek an appropriate legislative approach. 

However, most of the AI action so far is happening within federal agencies and in the scientific community. There’s potential for the technology to not only boost biomedical research but also improve diagnosis and the quality of patient care.

Administrative push: In October, President Joe Biden issued an executive order creating guidelines and support for the use of AI across the federal government. That included a directive to the appropriate agencies to “advance the responsible use of AI in healthcare and the development of affordable and life-saving drugs.”

Already, the National Institutes of Health is embracing the technology for a wide range of uses, including data analysis, medical imaging and research. 

And the Food and Drug Administration has authorized nearly 700 AI-enabled devices. The FDA contends that AI can strengthen its operational systems by processing and analyzing complex information faster, “including data from medical imaging or digital health technologies.”

Monica Lopez, a brain scientist who advises different industries on the use of AI, also said the technology could play a key role in improving clinical trials, which tend to be expensive, time-consuming and not inclusive. 

“So imagine if you can create simulations of this,” said Lopez, co-founder and CEO of Cognitive Insights for Artificial Intelligence. “It would be much faster, so much more efficient and reduce costs. It would allow for more to engage in that, so in a way, it democratizes access capabilities.”

The Spotlight Interview: Rep. Yvette Clarke

We spoke with Rep. Yvette Clarke (D-N.Y.), a member of the House’s artificial intelligence task force, about how Congress should approach AI’s opportunities for health care and medical research. 

Clarke embraces AI’s potential to revolutionize health care, diagnosis and treatment. Still, she wants to ensure the technology is using data that takes into consideration health care disparities for women and communities of color.

The Biden administration and patient advocates also share the concern about representative data in AI systems.

Clarke is the lead sponsor of the Algorithmic Accountability Act, a bill that would require some businesses that use AI-enabled systems to report on how the technology is impacting consumers, including those with a significant impact on the cost or availability of health care.

Here are key points from the conversation with Clarke. They have been edited for brevity and clarity.

“When you’re relying on the information that is provided through years and decades of medical research that has pretty much been based primarily on white male health, you’re in danger of continuing to promote health care, treatment and diagnoses that have not ever been a good fit for those who don’t fit the profile.”

“Part of the challenge that we face right now as a civil society is really catching up in terms of medical research and data that would then inform our datasets, which then go into the work that AI can do on our behalf.”

“We do a lot of investment as a nation in research.”

“Given the role that NIH and CDC play in discovery and research, I think that we have an opportunity in influencing the federal enterprise to serve as a repository of information that can revolutionize the use of AI in a way…that is reflective of how different segments of the population are receiving health care.”

“At the end of the day, it’s our obligation to leave room for innovation and make sure that we are not stifling that in the work that we do.”

“There are a whole host of ethical questions. There are HIPAA laws that we know need to be maintained in an environment where we have adversaries looking to exploit information and data…to misinform and miseducate.

“These are all compounding the sort of careful way in which we have to go about looking at how we maximize on AI in health care and in the medical space.”

The Policy Pipeline

There’s been a flurry of activity around AI and health care in Congress in recent months as lawmakers plunge into potential policymaking.

House Energy and Commerce Committee Chair Cathy McMorris Rodgers (R-Wash.) has called for a national data privacy standard as the starting point for AI health care legislation. McMorris Rodgers and Senate Commerce Committee Chair Maria Cantwell (D-Wash.) introduced bipartisan data privacy legislation in April.

Some individual members interested in AI have their own proposals. For example, Rep. David Schweikert (R-Ariz.) has introduced the Healthy Technology Act, a bill that would qualify AI and machine learning technologies as practitioners eligible to prescribe drugs.

Schweikert, a House Ways and Means Committee member, also views AI as a tool that could help address the growing U.S. debt by shrinking health care costs while also keeping people healthier.

There are examples “where you can remove costs of the system and make it much faster and much cheaper, and they’re already in progress,” Schweikert said of AI opportunities.

When you’re talking about matters that are going to hurt people — in this case users of the health care system — you’ve got to have guardrails, you’ve got to have protections to ensure it.

Sen. Ron Wyden (D-Ore.)

New tech, new rules: Some Democrats want to create guardrails for AI now, when the health care sector is increasingly adopting the technology. Senate Finance Committee Chair Ron Wyden (D-Ore.) introduced that chamber’s version of Clarke’s bill addressing algorithmic bias in health care data.

Wyden told us that AI’s novel qualities warrant new rules and a unique approach, especially if it’s to prevent discrimination in the health care field.

Wyden also sees an opportunity for policymakers to bolster the use of AI to help rural communities access health care, addressing a critical gap.

“When you’re talking about matters that are going to hurt people — in this case users of the health care system — you’ve got to have guardrails, you’ve got to have protections to ensure it,” Wyden said. “At the same time, if you’re talking about the general deal in terms of AI, you do need a wide berth to encourage all of these kinds of approaches.”

In reality, most lawmakers are still in the early stages of learning about the complexities of AI. That means it may be a while before definitive AI legislation moves through Congress.

— Laura Weiss


How AI supports tuberculosis detection


Tuberculosis (TB) infects 10 million people per year. Diagnosing TB early is important, but can be difficult because its symptoms can mimic those of common respiratory diseases. While chest X-rays are a common and cost-effective way to screen for TB, there is limited access to experts that are able to interpret results. That’s why Google is partnering with Apollo Radiology International in India to develop an AI system to interpret chest X-ray scans for early signs of TB. This collaboration will also provide millions of free screenings over the next 10 years.


Learn more

Also in this series:



We examine how the U.S. is embracing AI as a tool to bolster its cyber defenses and what Congress is doing on the issue.

Read more

Unlock access for more

Sign up to receive our free morning edition every week day, and you'll never miss a scoop.

Already subscribed? Sign in.