Skip to main content
Healthy, senior man running with group of friends Healthy, senior man running with group of friends

Ensuring responsible use of AI in health care

Discover how we responsibly use artificial intelligence and machine learning solutions to help people live healthier lives and make the health system work better for everyone.

 

What is responsible AI?

If we think of artificial intelligence (AI) as computer technology simulating human intelligence, then we think of responsible AI as the practice of ensuring that AI is designed and used fairly, ethically and safely.

Our Responsible Use of Artificial Intelligence (RUAI) program is inclusive of mechanisms and practices that ensure that AL/ML solutions are designed, developed and maintained following consistent, established standards.

Our guiding principles

The Responsible Use of AI Guiding Principles provide the foundation for all AI/ML work conducted across the UnitedHealth Group enterprise.

Reliability & Quality

We believe AI/ML Solutions should perform reliably. We will employ methods to test and monitor integrity, reliability and performance, with a focus on oversight, designed to ensure that AI/ML Solutions meet quality expectations and fulfill their intended uses.

Fairness

We believe AI/ML Solutions should treat people fairly. We are committed to assessing the performance of our AI/ML Solutions for potential bias and disparate impact on vulnerable groups and those who may be at risk of exclusion from the health system.

Accountability

We are accountable for the AI/ML Solutions we implement. We will establish and execute processes, governance and monitoring of our AI/ML Solutions that will enable swift remediation of emerging issues or unforeseen adverse outcomes. AI/ML Solutions will not replace clinical judgement.

Transparency

We believe that transparency and explainability enable the responsible use of AI/ML. We will monitor our use of AI/ML across the enterprise to understand how our AI/ML Solutions are used and how they operate.

Privacy & Security

We will safeguard data privacy and security interests in the design, deployment, and use of AI/ML Solutions. We will incorporate our organization’s rigorous, longstanding privacy and security practices consistent with applicable laws, regulations and policy requirements.

Continuous improvement

We recognize the rapidly evolving AI/ML landscape and will continuously improve our processes, including the application of industry best practices and regulatory guidance.

"

Intelligently improving human experiences

We continually assess and improve our practices based on our own learnings, industry best practices and regulatory guidance. Through our AI/ML governance, we assess intended uses, the appropriate use of data, performance, fairness and bias testing results and provide ongoing clinical, legal, compliance, business and technical oversight.

The responsible use of artificial intelligence and machine learning technologies is essential to the future of health care. Our responsible AI/ML processes, practices and principles are initiated prior to model development and are vital throughout the entire lifecycle of each of these tools. Mitigating potential bias in AI/ML models is a continuous improvement effort, not only because it makes the deployment of AI/ML safer and more impactful, but also because it can help to realize more equitable outcomes. The responsible use of these technologies is the responsibility of everyone in our enterprise — not just our AI/ML scientists — and should be at the forefront of execution across the industry so we can all make the health care system better for everyone.

Chief AI Officer, UnitedHealth Group

Rahul Bhotika

Video

See how using AI responsibly helps people live healthier lives

From improving the accuracy of a diagnosis to better predicting a patient’s response to treatment to managing care faster and more cost effectively, AI is already helping us make the health care system work better for everyone. See how AI was used to revolutionize the utilization review process.

"
"

Increased fairness, accuracy, and safety; reduced risk

We use AI/ML as inputs in the decision-making process and combine these with inputs from patients, providers, experts and partners, thereby helping to mitigate risks related to fairness, accuracy and safety, consistent with our guiding principles.

A review board of cross-functional expertise

Our AI/ML review board, the governance body for the RUAI Program, is comprised of senior leaders with cross-functional representation — including clinicians and clinical ethicists, business and operations leaders with analytics, technical, legal, compliance, regulatory and privacy expertise.

"