A new study from Innovaccer found that nearly 80% of payers now prefer deploying vendor-built artificial intelligence tools over developing internal capabilities.
The study draws insights from leaders of 63 health payer organizations, ranging from local health plans to national payers, the healthcare technology and AI company said in a press release. Respondents were surveyed from mid-December to mid-January 2025 and included senior executives and executives.
Innovaccer CEO and co-founder Abhinav Shashank told Fierce Healthcare that the move to outsourced solutions reflects a focus on how to “truly operationalize AI.”
“What we’re seeing is the emergence of a question of how can enterprises provide an effective platform that enables more agential orchestration,” Shashank said. “The reality is that this technology is a great addition to the way payers operate.”
The use of AI is rapidly increasing among payers as well as patients and healthcare providers. Innovaccer’s survey found that nearly 78% of respondents reported using solutions to improve care, and three-quarters of respondents reported that they were “actively pursuing or gradually experimenting” with AI in care innovation.
Planned investments further underscore this shift, with 75% of respondents reporting plans to spend an average of $10 million on AI over the next three to five years. One-third of respondents report having investment plans of at least $20 million.
However, 86% of respondents said they are not fully ready to deploy AI at scale. Shashank said the most important aspect of this is the “context and data infrastructure” behind the solution.
The study notes that payer data often resides in “legacy systems and silos,” which can limit the pace of AI adoption. The most commonly reported infrastructure barrier is interoperability, followed by real-time data access, poor data architecture, and cloud capabilities.
“Today, enterprises don’t necessarily have the core data and context infrastructure that allows them to operate these AI frameworks at scale,” he said, adding that enterprises need to “build out” the necessary infrastructure.
Additionally, according to Shashank, the industry’s readiness “relies heavily” on systems with the right data and context infrastructure.
“Without the data and context infrastructure, most of your AI efforts and investments won’t necessarily scale as a company,” he said.
Related:
Despite the barriers, 44% of respondents reported that they believe AI is essential to their organization’s member care goals, and 62% cited implementing AI to support member personalized navigation as a “use case critical to payer success” over the next three to five years.
“The fundamental shift that we are effectively starting to see is that people are not buying AI, they are using AI to solve problems,” Shashank said.
