Cold, Hard Math: The Rise of Artificial Intelligence in Insurance Claims Handling

By Steven J. Anglés, Attorney at Law

Recovering from a traumatic injury is often a slow process that unfolds in ways patients often can’t anticipate. The full extent of an injury, or the effectiveness of treatment might not be fully known for a long period of time, and only after the patient’s life has seen a roller coaster of personal and professional changes. The same can be said of navigating the insurance claims process. Coordinating insurance benefits, interpreting insurance contracts, and hoping for the end of health care, returning to full activities, and trying to find closure in an ongoing insurance claim can take years and test the patience of even the most reasonable people.

Recent technological changes are attempting to take all of these human variables and generate insurance claim outcomes in seconds, without involving humans at all. Over the past decades, insurance adjusters have increasingly used computer programs to settle personal injury claims in an attempt to save money and increase the speed of the claims process. But, the latest attempts to automate the claims process have gone much further by eliminating the need for claims adjusters through the use of artificial intelligence[1]. While this may seem like a reasonable approach for both insurance companies and patients, the use of artificial intelligence is likely to result in an extremely one-sided and arbitrary claims handling process that will undoubtedly reduce costs for insurance companies, while leaving injured patients with claims that aren’t evaluated in a meaningful way, or a way that fully takes into account how humans would respond to such a claim to resolve it.

In May of 2017, Zürich Insurance announced it was deploying artificial intelligence to evaluate personal injury claims after testing showed claims processing could take years’ worth of information and reduce it to a single decision in a matter of seconds. Insurance companies in Japan and Great Britain have begun doing the same, replacing staff members with sophisticated computer algorithms to perform the work previously done by humans.[2]  Computers using artificial intelligence, such as IBM’s “Watson” will be capable of reading medical documentation in order to collect information necessary for paying medical bills using factors such as medical histories, length of hospital stays, and surgical procedure names.  Cognitive technology could be used to review medical reports for evaluating personal injury claims-reducing review time of 10-100 page documents from 58 minutes down to five seconds.[3]  The system would also be able to check insurance claims against their corresponding insurance contracts to determine whether there are any coverage problems that need to be raised against customers.[4]

Large insurance companies are not the only businesses interested in the idea of expanding the use of artificial intelligence in the processing of claims.  There are an estimated 2,000 start-up companies centered around artificial intelligence, many of which are developing software and apps in order to replace human actuaries, reduce labor costs, and allegedly make more accurate predictions.[5]  Lemonade, for example, is a property and casualty insurance company funded by corporations such as Google that touts it can insure homeowners and renters in 90 seconds and pay out claims in 3 minutes through an app using a smart phone.[6]  Tractable is a London-based artificial intelligence corporation that promotes its ability to streamline automobile property damage repair payments by comparing uploaded photographs and repair estimates, reducing the cycle time from days and weeks to minutes, without involving a human claims handler.[7]  Yet another startup company, RightIndem, looks to put the injured party in charge of the entire claims process for motor vehicle collisions by inviting customers to make and process claims at “their pace” using their smart phone, tablet, or personal computer.[8]  The one constant among all of these companies is the promise to reduce claims “leakage,” a term used to describe the dollars lost through claims management “inefficiencies” that raise costs and lower profits for insurance companies.

While the use of artificial intelligence to process personal injury claims has been described as a “focused set of initiatives to improve customer experience and… effectiveness in [the] claims” function,[9] serious doubts remain about the ability of artificial intelligence to appreciate the very human impact, and unique nature, of each traumatic injury on the individual.  This artificial process also cannot replace or replicate the human behaviors, emotions and experiences that impact a judge or jury tasked with evaluating a personal injury claim in court. Even insurance adjusters at times have been forced to acknowledge the inherent limitations of mathematical formulas and computers in successfully evaluating and resolving personal injury claims.  In short, artificial intelligence is unlikely to fully appreciate the nuances that are part of each and every personal injury claim, and unique to each patient.  The attorneys and staff at the law firm of Adler Giersch remain committed to ensuring that our clients are never seen as just names and numbers on a piece of paper.  If you, your staff, or your patients have any questions regarding the processing of insurance benefits, feel free to reach out to us.

 


[1] “Artificial intelligence”, or, “machine learning”, is the theory and development of a computer system able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translating between languages. Oxford Dictionaries.

[2] An”algorithm” is a mathematical instruction that allows a computer to process information. As more and more algorithms are calculated, the computer can begin to learn on its own and propose solutions. Deangelis, Stephen F. “Artificial Intelligence: How Algorithms Make Systems Smart.” Wired, September 2014.

[3] “AI and Insurance: Are Claims Jobs in Danger?” Management, January 9, 2017.

[4] http://mainichi.jp/english/articles/20161230/p2a/00m/0na/005000c

[5] Nanalyze. “10 Artificial Intelligence Startups in Insurance.” Nanalyze, May 21, 2017.

[6] https://www.lemonade.com/faq#policy

[7] https://tractable.ai/products/insurance

[8] https://rightindem.com/

[9] “AI and Insurance: Are Claims Jobs in Danger?” Management, January 9, 2017.