Prudential will tap Google's MedLM gen AI models to verify medical claims

Prudential is turning to generative artificial intelligence (gen AI) to verify medical insurance claims in an effort to speed up processes. 

 The insurance company announced, in a statement on Thursday that the move also aims to cut the risk of errors caused by manual data entry. Prudential will tap Google’s MedLM family of large language models that have been trained for healthcare use cases.

Google unveiled the model last December, which evolved from predecessors Med-PaLM and Med-PaLM2. The US tech giant said MedLM offers two models built on Med-PaLM 2 to handle basic and complex workflows

Also: Asking medical questions through MyChart? Your doctor may let AI respond. 

The first (and larger) MedLM model is designed for complex tasks, while the second (medium model) can be tweaked and scaled across tasks, Google stated. 

Prudential hopes it will better support human-made decisions, improving the accuracy and efficiency of medical insurance claims. The insurance company reiterated that the AI model typically is used by healthcare organizations to develop applications for doctors, hospitals, and other medical providers to enhance patients’ and healthcare workers’ experiences. Furthermore, Prudential is among the first insurance providers globally to use MedLM. 

The company will begin its use of MedLM in Singapore and Malaysia for three to four months, during which, Prudential will apply the gen AI model to select medical insurance claims. This will involve analyzing and pulling relevant information from documents submitted with claims, such as diagnostic reports and prescriptions. 

Also: Hooking up generative AI to medical data improved usefulness for doctors

The use of MedLM here will mitigate the potential for manual errors and enable claims to be processed faster and with higher accuracy, Prudential said. 

It will then compare analyses and advice powered by the gen AI model against decisions based on current approval processes during the initial rollout period. This will allow Prudential to identify areas where MedLM can facilitate optimal productivity gains and impactful advice — while maintaining a “human in the loop” at critical stages of the decision-making process, the insurance company added. 

In addition, proof-of-concept tests were carried out before the rollout, with MedLM doubling the automation rate of claim assessments and improving the accuracy of claim decisions. These allowed for higher volume and speed of claims processing, it said.

“Prudential’s early tests with MedLM demonstrate that gen AI can play a major role in efficiently tackling the growing volume of health insurance claims, resulting in more frictionless processing and a faster turnaround time for customers,” said Arjan Toor, CEO of Health at Prudential. 

Also: How GenAI got much better at medical questions – thanks to RAG

“In a fragmented and often confusing healthcare landscape, data and AI are enabling us to provide care beyond coverage for our customers,” said Toor, who is based in Singapore. The partnership with Google Cloud will provide Prudential with a “first-mover advantage” in adopting Gen AI to improve customer experience “at an important moment of truth,” he said. 

“This is just the first step in using Gen AI to deliver seamless, digitally enabled healthcare experiences at every step of our customers’ health journey — from the point of diagnosis, through treatment, recovery, and prevention,” he added.