2026 Encoding Health Equity Summit - Opening Keynote: Choices: Building Ethical AI Systems-2026 Encoding Health Equity Summit Opening Keynote

Video Transcription

Video Summary

The speaker argues that AI is a major turning point for healthcare and society, and that professionals must actively shape it rather than let it shape them. Drawing on the history of healthcare quality, electricity, and other general-purpose technologies, he explains that new technologies often arrive with fear, greed, and inequity, but can still produce major gains over time if guided well.<br /><br />He says AI is unusually powerful, widely desired by doctors and patients, and could help reduce healthcare costs and lessen the field’s reliance on labor alone. However, AI also carries serious risks: it can be trained on unrepresentative data, reproduce old biases, use harmful proxies like cost instead of medical need, and deepen disparities in access. He gives examples of biased sepsis models, race-related misinformation in language models, and algorithms that favored white patients because they used healthcare more.<br /><br />The speaker emphasizes that outcomes depend on choices about training data, oversight, policy, and implementation. He highlights positive examples such as YouTube privileging scientific sources during COVID, the HITECH Act expanding electronic health records, and targeted AI use to reduce no-show rates for underserved patients.<br /><br />His final message: use AI, join governance efforts, and use practical equity tools so the technology helps scale justice, fairness, and better health for all.

Keywords

AI in healthcare

health equity

algorithmic bias

general-purpose technology

healthcare quality

biased data

medical governance

digital health