## Summary

The article discusses the impact of biased AI in the healthcare and life sciences industry, highlighting the disparities in access to care and the use of generative AI by historically underserved communities. It emphasizes the need for responsible AI governance to ensure fairness, transparency, and accountability in AI models. Drawing parallels to the evolution of the US Food and Drug Administration, the article proposes five key steps for ensuring equitable outcomes from generative AI in healthcare and life sciences.

## Delivering Ethical AI in Healthcare and Life Sciences

The COVID-19 pandemic brought attention to health inequities, especially affecting historically underserved communities. The National Institute for Health (NIH) reported that Black Americans were disproportionately affected by COVID-19 due to factors such as limited access to care and underlying health conditions.

Generative AI, like ChatGPT, has been found to provide inaccurate and even dangerous medical advice. This raises concerns about the potential harm faced by communities relying on such technology instead of professional medical assistance.

To proactively invest in ethical AI for equitable outcomes, the article suggests the need for AI governance, trust, security, and regulatory considerations. It also emphasizes the importance of institutional innovation and draws a comparison to the historical evolution of the US Food and Drug Administration.

## Institutional Innovation and Equitable AI Outcomes

Institutional innovation is required to ensure equitable outcomes from AI, similar to the regulatory changes following the Elixir Sulfanilamide disaster, emphasizing the need for responsible AI governance in the healthcare and life sciences industry.

The article outlines five key steps to ensure that generative AI supports vulnerable populations, including operationalizing principles for trust and transparency, appointing individuals for accountability, empowering domain experts to curate trusted data sources, mandating auditable and explainable outputs, and requiring transparency in AI integration.

The authors argue that by learning from the historical evolution of regulatory bodies like the FDA, it is possible to institute changes that make AI more reflective of the communities it serves and earn people’s trust.

### FAQ

1. What is the significance of the NIH report mentioned in the article?
The NIH report highlighted the disparities in COVID-19 rates among Black and White Americans, attributing them to limited access to care, public policy inadequacies, and a higher burden of comorbidities.

2. Why is responsible AI governance important in healthcare and life sciences?
Responsible AI governance is crucial to ensure fairness, transparency, and accountability in AI models, particularly in providing medical advice and services.

3. How can institutions innovate to ensure equitable outcomes from AI?
Institutional innovation involves implementing systemic changes to make AI more reflective of the communities it serves, drawing parallels to historical regulatory changes in the healthcare industry.

4. What are the key steps outlined to ensure generative AI supports vulnerable populations?
The key steps include operationalizing principles for trust and transparency, appointing individuals for accountability, empowering domain experts to curate trusted data sources, mandating auditable and explainable outputs, and requiring transparency in AI integration.