CVS Health leverages AI for personalized care and reduced administrative burdens

Nikolas Kokovlis | Nurphoto | Getty Images

This week was AHIP 2026 in Las Vegas, the annual conference hosted by the nation's leading trade association for health insurers.

The two-day event featured discussions spanning the healthcare landscape, from new industry commitments on prior authorization and the growing role of AI to ongoing debates over how to balance pharmaceutical innovation with affordability.

CVS Health spoke to CNBC about how the company is using AI and other technologies to reduce administrative burdens and deliver more personalized, proactive care experiences.

Below are some highlights from the conversation with Katerina Guerraz, Aetna's chief operating officer and president of Medicaid, and Tilak Mandadi, CVS Health's chief experience and technology officer.

Real-time data sharing could be a gamechanger

Guerraz said some of the healthcare system's biggest pain points could become easier to address over the next several years as insurers and providers work to reduce administrative burdens and share data more seamlessly.

She pointed to the industry's existing efforts to streamline prior authorization and other insurance processes, but said broader "interoperability" between payers, providers and other healthcare stakeholders could be the "bigger thing that's going to transform this industry" in the next three-to-five years.

"When anyone who's in the health ecosystem starts sharing data in real time, that actually changes everything," Guerraz told CNBC. "You're not going through all these different intermediaries of people who are processing different kinds of information."

That could help patients access care faster, she added.

Conversational AI tool could help patients

CVS is betting that its conversational AI assistant, which launched in the fall on Aetna's website and mobile app, can help members navigate the health-care system more easily by answering questions, anticipating needs and reducing the need for customer service calls.

The assistant is designed to go beyond the capabilities of traditional chatbots and automated phone systems, Mandadi said. Rather than simply responding to a member's question, the tool aims to understand what the person is trying to accomplish and guide them through the next steps. For example, a member asking whether a surgery is covered could also receive information about expected out-of-pocket costs, prior authorization requirements and in-network care options, he added.

"We know why people call us most of the time. So how do we solve those and get them answers before they even have to reach out?" Mandadi said. "But then if they do, the conversational AI is the one who can get it faster."

He said it's a proactive approach that gives members information before they even have to pick up the phone – and the company benefits, too.

Mandadi said the company's pharmacy business previously received more than 500 million calls annually, many of which involved simple questions such as the status of a prescription. But today, roughly 75% of those interactions are fully resolved through conversational AI, allowing pharmacists and customer service employees to focus on more complex issues, he said.

Personal AI assistant could anticipate patient needs

CVS also envisions a future where patients can have a "persistent" personal AI assistant that actively manages and anticipates their needs in real time, Mandadi said.

That's one key feature of CVS and Google Cloud's Health100 platform, which is set to launch this year.

The AI assistant would sit across a member's insurance, pharmacy, and clinical data and guide health decisions over time, he said. That tool goes beyond today's chatbots: Mandadi described a consent-based system that maintains a continuous view of a member's health history, conditions, medications and benefits and uses that information to anticipate needs rather than simply respond to questions.

It's another example of shifting health-care interactions from reactive to proactive, Mandadi said. Instead of waiting for members to call with questions or miss key steps in their care, the assistant would flag issues in advance and suggest follow-ups and recommendations in real time.

He said the goal is to reduce the "healthcare homework" patients often have to navigate on their own, such as understanding benefits, managing chronic conditions or keeping track of appointments and medications.

The assistant could even detect early warning signs, such as missed medication refills or worsening health indicators, and intervene before they escalate into more serious issues.

He said the system would eventually integrate multiple data sources, including claims data, pharmacy records, electronic health records and potentially wearable devices, if patients give their consent.

Mandadi said that the assistant could ultimately help patients reach better health outcomes and reduce their out-of-pocket costs.

"That's the real beauty of it."