Craig Kurtzweil, chief data and analytics officer for UnitedHealthcare's commercial business, shared insights with Digital Insurance related to how the company is using AI and advanced tech.
How is UnitedHealthcare using AI and advanced technology?
UnitedHealthcare is focused on helping to reduce friction and complexity to enhance the healthcare experience for our members and employer clients. We are leaning into the
Our newest tools are helping us to streamline processes, remove administrative burden, and lower costs for members.
One very important place we are leveraging AI is in UnitedHealthcare's
Beyond Surest, we have more than 1,000 AI use cases in production across UnitedHealth Group, including many within our commercial business. In our commercial business, we are able to uncover trends that may help employers design more effective benefits strategies or help our advocates more efficiently and effectively engage members to help answer their benefit and coverage questions and better meet their needs.
Our Claims Assistant AI tool is one of the customer-facing digital enhancements rolled out this year to make reimbursement for out-of-network claims simpler for our members. With its prompt engineering capabilities, this AI assistant simplifies the process by automatically populating all the medical field values, making reimbursement submission faster and easier than ever.
Much of our use of generative AI is focused on cautious design and implementation of solutions that make information more accessible to members, helping streamline administrative processes. To make information more accessible, we've introduced new tools to help connect members with needed care like Smart Choice, a provider search that leverages AI to help match members in a more personalized way.
AI solutions like these are helping us to meet consumer expectations for access to personalized, simple data-driven information to help support better decision making, even from the palm of their hands.
What tools are being used in customer-facing situations and how are they being received?
We have seen success in reducing administrative burdens for members with our Claims Assistant AI tool which has helped to cut the time of submitting out-of-network claims from 12-15 minutes down to three minutes.
And our provider search tools are being used at scale to simplify care navigation: 18 million AI-enabled searches to help customers find a doctor were conducted in the first quarter of 2025. While Members Like You, which is one of our newer digital tools, has recorded more than 40,000 interactions so far.
Digital tools are also helping us to get our members the assistance they require and resolve support issues more quickly. Last year, 65 million customer calls were initially answered by an AI chatbot. One of the tools helping to make information and non-clinical answers more accessible for members is our Conversational AI for Voice, a virtual assistant that can chat in real time and determine how best to provide support related to benefits and coverage questions through self-service or a live agent, getting members the information they need faster and with greater accuracy.
What tools are being used internally?
When a member reaches out with a health care question, they're connected to an advocate, and we've added AI tools to support those experiences so the advocate can focus more on the member. Advocate AI Tools focus on making non-clinical information more easily accessible. Examples include:
- Welcome experience: Provides key member history based on claims activity or previous calls to the advocate before the call to streamline authentication and personalize support.
- In-call assist: Uses voice analytics to understand the member's needs and provides benefits, coverage and cost answers from the member's benefit plan.
- In-call summarization: Summarizes the conversation for advocates to add to the member's file so that when they call back the advocate doesn't have to spend a long time understanding member call history.
We are also deploying generative AI solutions to support internal administrative tasks like Alfred, which was recently rolled out to help reduce time spent on administrative tasks such as goal setting and feedback.
