Envisioning Claims Management Nirvana
During our recent webinar on Augmenting Claims Management with Automation and AI, panelists from Sedgwick and Accenture were asked to envision the future state of claims management. What might this look like from the claims operation, the employee’s role, and the technology perspective?
The Claims Operation
Accenture led with a controversial 5 year vision - a world where there are no claims at all. Instead, once a policy has been executed on both sides - technologies such as IoT, Telematics, Blockchains AI and Automation then takeover to monitor, track and handle claims and communications.
Bringing the vision back to the next 2 - 3 years, panelists agreed that AI can minimize the number of claims an insurer has to deal with, so they can focus on making better decisions rather than following the “cow path”, a theory where the same decisions are made regardless of new information.
Kenneth Saldanha, a Senior Managing Director at Accenture agrees:
“...I think what we do with AI is on the one hand, really take out a huge part of this volume, to begin with, based on this shift over to risk mitigation, and then really start the leapfrog past a lot of this paved cow path that we have in the claims process right now to a very different approach of making decisions and interacting with the customers in a much more engaging manner”.
In an idealistic future, AI would also be able to lessen the time it takes for adjusters to collect information needed to settle claims.
Andrea Buhl, the President of Managed Care at Sedgwick believes insurers could leverage AI in the future by collecting and synthesizing pertinent information to streamline the claims management process.
"If insurers could delegate the more administrative tasks of claims adjusting to AI, this would allow their adjusters to work on the human decision-making that technology cannot".
Roots CEO and Co-Founder, Chaz Perera succinctly offers
“...this is, I think, where AI as opposed to traditional automation technology is going to be really valuable.”
Role of People
While AI will no doubt play a large part in the claims management of the future, it will not eliminate the need for human intervention. Humans have an important role in not only verifying the accuracy of AI but confirming whether the decision being made is the right one, especially in such a highly regulated industry. Chaz says it best, that AI should be “supporting as opposed to eliminating the need for people.” It introduces complex technology in a controlled manner and allows the individuals responsible for making decisions to do so confidently.
In addition, the industry has been facing a talent crisis where a majority of the workforce is retiring without the younger generation replacing them. As a result, insurers may start to rely on AI and technology adoption to attract new talent. However, as Andrea mentions, “Companies are gonna have to integrate technology and AI into the processes in order to attract the younger generation, while at the same time, they're going to need to ensure that those folks that are currently in the labor market that are aging are going to be able to adopt that technology as well.” It’s an interesting challenge that companies will have to consider.
Technology
Insurers in the future may also adopt AI tools to support decisions and prediction. Chaz demonstrated Root Automation’s Digital Coworker, a large language model (LLM) system that can aid claims adjusters in making decisions. The digital co-worker can take unstructured data, such as an Accord form, and review the information, contextualize it, and then organize it for the adjuster. It can even make inferences on additional information that may be needed and send it back via API to a customer.
This type of technology can free up the claims adjusters to use their technical skill set to make decisions, rather than filtering through documents or other various types of datasets. It also streamlines the entire process and reduces the time it will take to settle claims, which can increase customer satisfaction.
However, AI does not come without its risks and challenges. The biggest challenge is that the systems are only as accurate as the developers who build them. Any information beyond what the AI systems were trained on will be assumed and not necessarily correct. This can have significant consequences, such as when it comes to making decisions on whether to pay or deny a claim.
There are also ethical considerations as well. Who is responsible when things go awry? Andrea provided an example of AI being used to diagnose or treat a patient. What if the data is incorrect? There are still many conversations to be had regarding how to manage this risk.
Current State
Of course, making the journey toward this vision cannot happen overnight. As a whole, the insurance industry is risk-averse and slow to adopt new technologies. Especially when it comes to introducing a new type of risk for insurers.
Adoption of Tech
Regarding the adoption of technology, Kenneth points out- “We used to talk about being at the intersection of business and technology. At this point, I think we would say that business is technology.” Insurance companies are going to need to adopt technology in order to remain competitive and stay relevant within the industry. But there can be significant challenges within an organization to incorporate this new technology within their processes. Especially if they have always relied on legacy systems or humans to complete certain tasks. Andrea states that if you rely solely on an operations team, it’s more difficult to visualize the larger picture. And IT teams don’t always understand the operations side. Insurers should find a mediator who knows both sides and has the dual skill set to incorporate technology into what is needed and already existing.
Removing Bias
Even the most honest and skilled employees will bring bias into their decisions. The adoption of technology allows insurers to remove or at least reduce the level of bias in decision making. The two biggest functions within an insurance carrier that are affected by this are the claims and underwriting departments. Chaz says, “There’s an aspect of our business that is still human intensive and we just need to make sure that we have the right humans involved in assessing risk at the enterprise level.” He suggests using AI technology as a second set of eyes so there are checks and balances. Eventually, this reinforced learning will hone the AI to produce the right outcomes for each company.
Recommendations
You can’t doubt the advancements AI can bring to an organization. But where should companies begin? Andrea recommends starting small. “You start with one area, and then you can continue to progress it. You can continue to refine it. It doesn't have to be a total rework of your entire organization to start implementing AI.” Kenneth adds that you must focus on a suite of capabilities that will create a material impact and make a difference in your company’s performance.
A digital co-worker can also be beneficial. While large language models such as ChatGPT are useful, they only have a 50-60% accuracy rate. Roots Automation’s digital co-worker has an accuracy of 98% and can do in seconds what may typically take an employee a half hour or more. By eliminating these linear tasks, insurers allow their adjusters to focus on the tasks that require critical thinking, ultimately optimizing their efficiency and getting us one step closer to the insurance nirvana we envision today.