Optum Rx is driving an AI-powered transformation at the leading edge of pharmacy care. And for us, the promise of AI is real – not aspirational. As we move forward, we’re using AI responsibly to support, not replace, critical human decision-making.
Explore how we’re using AI capabilities to improve the way we serve members and clients. This transformation is driven by AI’s unique ability to synthesize vast amounts of data and turn it into a personalized service experience.
Optum Assistant is a powerful AI-driven experience that:
It’s a platform-level innovation that turns complexity into clarity and friction into flow.
Optum Assistant is your always-on companion designed to simplify the experience for clients, members and our advocates – one intelligent conversation at a time.
Here’s an example of how it works:
The member wants to know if their medication can cost less. Tap to start, and you can follow along while Optum Assistant guides the member through the steps to understand their options quickly and automatically.
By putting the power of self-service in members’ hands, we can reduce effort and empower both consumers and our service agents. That drives meaningful improvements in the way our clients experience their benefit plan.
Optum Assistant is more than an assistant. It’s a unique, different way of engaging consumers, powered by Optum. Using AI and market-leading technology, our Optum Assistant:
We’re using AI to predict member needs in the future, including things the member might not even realize.
For the last 5 years, Optum Rx has been steadily developing and refining an advanced AI model that proactively identifies members at risk for service issues and resolves them. This often occurs before they’re even aware of them. This part of the system identifies potential member concerns related to drugs that are out of stock, and prescriptions that are nearing exhaustion or too soon to fill.
Here’s an example of how it works:
Ann is a regular Optum Rx home delivery customer who relies on monthly thyroid medication refills. Our predictive model detected that Ann was within six days of running out of her prescription and had not yet requested a renewal. We use AI to ensure that Ann will get her prescription refilled on time, even if she doesn’t request it or respond to automated prompts right away.
If Ann does respond to our prompt, the refill proceeds while Optum asks her if she would like to sign up for auto-refill. If she still does not respond, the refill proceeds as the technology works to move Ann to auto-refill.
The result is better service, but also avoiding hundreds of thousands of unnecessary calls: 80% of potential member issues identified by AI avoid callbacks or escalation.1
If someone does need to call, our system leverages the predictive capability of AI to anticipate why. Beginning with who the member is, the system automatically predicts the likely reason for their call, visit or chat – without needing to directly ask them.
By analyzing their recent activity, past claims and previous calls for example, we can predict the top potential reasons for the call. Then, we can use that prediction to personalize the experience or connect the member to the correct support, avoiding multiple transfers.
AI technology helps our members get smarter guidance and more personalized support. That means more moments where their questions are answered instantly and their experience feels effortless.
This isn’t just a technology story. It’s a human story about transforming the pharmacy care experience in a way that’s predictive, personalized and profoundly human for everyone.
Article
At Optum Rx, we’re using AI responsibly — fairly, safely, ethically — to make health care easier and better for the people we serve.
Article
Two AI experts share insights on how we’re using AI and advanced technology to improve everything from trend prediction to member services.
Article
From predicting drug trend to earlier fraud protection, see how we’re leveraging AI to improve operations for clients and members.
Sources
1. Optum Rx internal data, 2025.
