Elevance Health wants to use AI to simplify and personalize healthcare. Here's how the insurance giant is doing it

Artificial intelligence dominated the conversation at the Consumer Electronics Show (CES) in Las Vegas earlier this month.

While consumer-focused tech companies like Samsung, Google and Sony historically get the spotlight at CES, this year, Elevance Health had a big presence at the event, billed as the largest tech event in the world.

Elevance Health CEO Gail Boudreaux gave a keynote speech at CES about the use of technology to make healthcare more proactive, predictive and personalized. Elevance's strong presence at CES this year underscores its ongoing focus on consumer digital health as it continues to build out its member app, called Sydney.

The health insurance giant, which serves 117 million people, sees opportunities to leverage generative AI and other technologies to make healthcare easier to navigate, chief AI officer Shawn Wang told me during an interview at CES.

"We're looking at 'How do we get AI embedded in every major business?' We think about AI as being 'people-first.' You start with why. Why do we do this? What problem are we trying to solve?" Wang said.

Elevance, formerly Anthem and the nation’s second-largest insurer, is looking to use AI to empower consumers and make healthcare less complicated for patients, providers and its employees, Wang noted.

"We have all these key stakeholders so how do we use AI and digital technology to make their lives easier? And make the complications go away. They're just so much opportunity for us to simplify processes," he said.

Elevance Health is exploring how to use AI and generative AI to simplify the consumer experience and streamline back-end processes for employers.

Boudreaux touched on these same themes during her keynote speech.

"There's a huge opportunity inside of our enterprise, and inside of the business of healthcare, to use AI and other technology to simplify the healthcare experience," she said during an on-stage interview with Peter Lee, corporate vice president of Microsoft Research and Incubations. "We have thousands and thousands of pages of policy documents and claim rules that we have an opportunity to simplify those and get those in the hands of clinicians very quickly."

Back in December, at the Forbes Healthcare Summit, Boudreaux said Elevance Health uses its clinical data platform, called Health OS, and artificial intelligence to help providers close gaps in care and reduce burdensome paperwork. The insurer's goal is to break down data silos and integrate data on patients' physical, mental and social health into a longitudinal patient record within electronic health record (EHR) systems, Boudreaux said.

"As we think about the consumer, it's about simplification and personalization. And that's an area that healthcare has historically struggled with. I think that's where AI has a huge opportunity for us through chatbots and other things to answer questions in a very personal way... so that we know your history, we know what is important to you and then we can channel you to the right resources," Boudreaux said during the CES keynote.

Elevance can use generative AI to provide members with personalized care recommendations, help them understand their health benefits and connect them with the right healthcare provider, she noted. The company has embedded AI models into its Sydney app to provide members with an interactive chat feature to quickly get answers to questions and built out a Personalized Match feature to help members find an in-plan provider who meets their needs and preferences, Wang noted.

The company is exploring specific operations and processes to use generative AI such as summarizing and extracting information and generating predictive insights, Wang said.

"We look at it through a value lens and we look at feasibility so high impact on value and also high feasibility, that's where the sweet spot is. All of that is built on what we call responsible AI frameworks," he said.

Elevance's approach to "responsible AI" is based on principles around privacy, security, accountability, transparency and health equity, Wang noted.

"Health equity is a big deal. We run robust processes, all the way from the source data to identify potential biases to doing a health equity analysis, to make sure every step of the way that we don't introduce arbitrary additional bias to it. It's about building guardrails," he said.

The company also prioritizes accountability for the impact of its AI tools.

"Historically, the folks doing the science work and the engineers don't necessarily see the end. They don't understand potentially the consequences of the work and what it can lead to. We do a lot of education and set out policies and compliance training around making sure everybody involved in AI is very clear about the potential implications and outcomes and making sure it's your accountability to make sure things are going in the right way and that you stay within that responsible AI framework and follow the policies and procedures we've identified," Wang said.

Transparency and "explainability" are critical to accelerate the adoption of AI in healthcare, he added.

"It cannot be a black box," Wang said. "We have to document every step of the way around how the data is being used, how the model is being developed, all the way to how things are being done in production and monitoring the process. And, most importantly, what's the impact to our consumers? It's people first," he said.

But AI development and adoption in healthcare will "move at the speed of trust," Boudreaux said during her speech.

Building trust gets back to the core principles around a responsible AI approach with a focus on transparency, accountability, equity and security, she noted. And, it's important to have humans in the loop to test out the models and clinicians have to be involved in any clinical decisions, she added.

"We have to get our own employees and everyone in the system to trust it along with consumers. We're trying to do that and we call it democratizing AI. We see opportunities with our associates to work on [Microsoft] Fabric inside of the company," Boudreaux said. Microsoft Fabric is an end-to-end analytics platform. "They test it, they get comfortable with it, they understand it but it's within a very controlled environment, so we're careful that it doesn't have an adverse impact," Boudreaux said.

Wang has led AI and data science teams at Elevance for eight years and his previous tech experience includes leading enterprise analytics at Sears. As a tech leader, Wang said he's excited about continued innovation with generative AI for the potential to empower users.

"Historically, you think about AI and digital technology, it feels like a black box. It's difficult for people to really understand how to use the technology. I do think generative AI is a major breakthrough in terms of it just being so easy to use. We empower our associates to use it," he said. "They're actually creating the solutions to drive the business forward. It's no longer just a group of AI scientists in the backroom. We're actually on the front line empowering our teams to translate their knowledge and intelligence and program that into the AI tools. That fundamentally transforms how AI is actually being ideated, being developed and being incorporated into the broader enterprise and it's a cultural shift."

There are ongoing concerns about the reliability and accuracy of large language models, and their capacity to "hallucinate" but developing a robust, responsible AI framework helps to tackle those issues, Wang noted.

The rapid pace of generative AI's ongoing evolution is a key concern, he said. "Things are moving really fast. We have to learn and there are new technologies coming out every day. However, we have limited bandwidth to learn these new things and figure out how to incorporate them into the enterprise. And how do we move fast to deliver things at scale? That's a continuing challenge. How do you scale? How do you get to truly enterprise impact? I think that's the number one topic for every technology leader at a large enterprise," he said.

Looking ahead at the next few years, generative AI has the potential to help organize and find insights in unstructured clinical data in medical records, Wang noted. The technology also will help to arm providers with better data tools and help simplify communication between all the different healthcare players.

"It has to be people first. You have to understand where the pain points are. This is a pretty typical supply and demand problem. We have limited resources serving a lot of people so how do you make sure these resources are truly empowered?" Wang noted.