As artificial intelligence rapidly transforms the American health care system, the APA today called on Congress to ensure that a deep understanding of human psychology is central to the technology’s development and oversight.
Before the House Energy and Commerce Subcommittee on Health, APA Senior Director for Health Care Innovation Vaile Wright, PhD, testified (PDF, 285KB)opens in new window that while AI holds immense promise, it also presents significant perils that require immediate legislative action. The hearing, “Examining Opportunities to Advance American Health Care through the Use of Artificial Intelligence Technologies,” featured a panel of experts from health care, technology, and academia.
“AI is not merely a technological advancement; it is a tool built by humans to be integrated into human systems,” Wright stated in her testimony. “Therefore, a deep understanding of human cognition, behavior, emotion, and interaction must be central to its deployment to ensure it serves patients and other users effectively, ethically, and equitably.”
Wright highlighted the potential benefits of AI, such as expanding access to care in the face of a behavioral health workforce shortage and alleviating administrative burdens that lead to provider burnout. However, she cautioned that public trust is fragile, with 60% of Americans uncomfortable with AI being used in their own health care.
This discomfort is not unfounded, Wright explained, pointing to two major risks:
- Health disparities: AI risks amplifying existing health inequities. For example, one widely used algorithm measured a patient’s illness based on their health care costs, unfairly attributing lower risk scores to populations that have historically spent less on care due to systemic factors, ultimately worsening health disparities.
- Unregulated products: The direct-to-consumer market is flooded with unregulated chatbots making deceptive and dangerous claims. Wright cited the example of an entertainment chatbot presenting itself as a “psychologist” that engaged in millions of chats and, in one case, appeared to validate a user’s violent thoughts. This is unacceptable, which is why APA has formally requested investigations by the Federal Trade Commission and the Consumer Product Safety Commission.
To realize AI’s promise while protecting patients, Wright urged the subcommittee to advance legislation and oversight built on a foundation of ethics, equity, and evidence. The APA’s recommendations include:
- Establish clear regulatory guardrails to ensure safety and efficacy, including prohibiting the misrepresentation of AI as licensed professionals and mandating human oversight.
- Protect vulnerable populations, especially youth, by requiring age-appropriate safeguards and robust data protections for adolescents, who are in a critical developmental period.
- Prioritize equity and mitigate harm by requiring that AI models undergo rigorous, independent testing for harms across diverse populations before they are deployed.
- Invest in research and AI literacy through significant federal funding to understand AI’s impacts, paired with comprehensive education for the public and providers.
- Enact comprehensive data privacy legislation establishing a right to “mental privacy” by safeguarding biometric and neural data that can be used to infer an individual’s mental state.
“Ultimately, we must ensure a human remains in the loop,” Wright concluded. “AI should be seen as a tool to augment, not replace, the clinical judgment and therapeutic relationship that are the bedrock of quality health care.”
link