In November 2023, the Subcommittee on Primary Health and Retirement Security under the Senate Committee on Health, Education, Labor and Pensions held a hearing to examine the implications of artificial intelligence (AI) in health care. The hearing brought together lawmakers, technology experts, health care professionals, ethicists, and patient advocates to evaluate both the opportunities and challenges associated with AI’s integration into health care delivery.

Key Findings

• Potential for Enhanced Care: AI tools demonstrate significant promise in improving clinical decision-making, streamlining administrative functions, and increasing access to quality care, particularly for underserved populations.
• Risks of Inaccuracy and Bias: There are concerns regarding AI’s potential to perpetuate or worsen existing biases in health data, possibly resulting in inaccurate diagnoses or unequal care for marginalized groups.
• Transparency and Explainability: The need for transparent and explainable AI models was emphasized to ensure clinicians and patients can understand the rationale behind AI-driven decisions.
• Data Privacy and Security: Witnesses highlighted the critical importance of safeguarding sensitive patient data, stressing the necessity for strong protections and compliance with privacy regulations.
• Regulatory Oversight: The Subcommittee discussed the need for regulatory frameworks that evolve with technological innovation while still promoting beneficial advancements.
• Workforce Impact: Testimony addressed how AI could affect the health care workforce, underlining the importance of ongoing training and adaptation for clinicians.
• Patient Involvement: Including patient perspectives in the design, deployment, and oversight of AI systems was recognized as essential for building trust and ensuring effectiveness.

Recommendations

• Develop and implement rigorous standards for AI transparency, accuracy, and fairness in health care applications.
• Increase federal investments in research to address data bias, model validation, and the safety of AI-enabled clinical tools.
• Establish clear guidelines for the ethical use of patient data, ensuring robust privacy protections and informed consent.
• Promote multidisciplinary collaboration among technologists, clinicians, ethicists, and patient representatives in AI development and deployment.
• Encourage ongoing education and training programs for health care professionals to keep pace with AI’s evolving role in practice.
• Strengthen oversight mechanisms to monitor AI systems’ real-world performance and impact, including pathways for reporting and addressing adverse outcomes.

Conclusion

The Subcommittee concluded that AI holds transformative potential for health care, but realizing these benefits requires thoughtful policy design, vigilant oversight, and unwavering commitment to equity, transparency, and patient safety. Legislative strategies will continue to be developed to foster innovation while prioritizing the interests of patients and providers.

AI Use and Associated Concerns

• AI tools can record and summarize doctor-patient conversations, with physicians editing AI-generated summaries, allowing the AI to learn from these edits.
• Concerns include the necessity of physician oversight, potential errors in coverage decisions, and the challenge of understanding how AI produces its outputs.