HAP's Latest News

5 Takeaways on the Future of AI in Health Care

Lawmakers weigh potential, safety, and ethics of emerging technologies

November 29, 2023

Federal lawmakers are exploring critical questions about the future of artificial intelligence (AI) in health care.

This morning, the House Energy and Commerce Health subcommittee hosted a hearing focused on the potential for AI to revolutionize health care. The hearing included key questions about how technology could reduce administrative burdens, while noting the need for proper guardrails to protect patient safety and data.

“As this technology continues to explode onto the scene, it has become especially prevalent in health care,” John Joyce (R-13) said during the hearing. “But like many industries where AI is seeing a dramatic increase in usage, there are and there will be certain risks associated with it that we must contend with as policymakers."

Here are a few key takeaways from the hearing:

New use cases: AI-driven systems can assist with provider documentation in emergency rooms, moving clinicians away from being “data entry analysts” and allowing more time with patients, said Dr. Michael J. Schlosser, vice president, care transformation and innovation, HCA Healthcare.
- AI also can help organize and process information in medical records to support the nurse handoff process, as well as staffing and scheduling to ensure providers have “the right team, in the right place at the right time,” he said.
Generative AI: Algorithms that use data to create text, speech, and other content are improving, but health care will need experts in AI and other “generative” technologies to properly assess these tools.
- The creation of a new AI tool does not always equate to better care, so providers must carefully assess implementation of new technology to avoid the “AI hype cycle,” Dr. Christopher Longhurst, MD, chief medical officer, chief digital officer and association dean, UC San Diego Health, said during his testimony.
Emphasis on data: “Fundamental data challenges” can lead new technologies to reach the wrong conclusions, said David E. Newman-Toker, MD PhD, Johns Hopkins University School of Medicine. “Gold-standard” data sets are needed to ensure these systems are trained to support clinicians and care.
- “AI systems that learn on faulty data will generally make the same mistakes that humans make,” he said.
Do no harm: Lawmakers noted concerns about human biases that could be embedded within new technologies, potentially affecting efforts to improve health equity.
- Transparency to patients, protecting sensitive data, and ensuring a human is always “in the loop” were among key concerns.
Next up: Lawmakers are considering how this technology should be regulated to support innovation while protecting patient safety and privacy, as well as proper reimbursement for this care. They also discussed the role of AI in insurance coverage decisions and medical liability.