Protecting Policyholders as Ai Is Developed for Insurance Claims Handling: Ensuring "decency and Humanity" in the Digital Age

• Publication year: 2024
• Citation: Vol. 4 No. 2

[Page 111]

Marshall Gilinsky and Madison Marlow ^*

Abstract: The integration of artificial intelligence (AI) within the insurance industry raises concerns that insurance companies might use the technology to unfairly curtail or deny policyholders' claims. Drawing on the historical example of the Colossus software, this article outlines the potential consequences of diminished human oversight in Al-driven claims handling. In the past, technology was used to boost insurance companies' bottom lines while undervaluing policyholders' claims. We may be seeing a similar situation unfold in real time with recent investigations into and lawsuits against certain health insurance companies for their alleged algorithm-driven claim denials. This article highlights the need for watchdogs and regulators to demand that AI tools under development afford "explainability" and protect policyholder rights. Insurance companies must stand by their fundamental duty of good faith to policyholders, and courts must maintain long-standing precedent that demands "decency and humanity" in insurance company claims operations.

Introduction

Technology rules the day, and it is being adopted across the insurance industry. Functions such as risk assessment, claims processing, and customer service are increasingly powered by artificial intelligence (AI). While AI has great potential for efficiency gains

[Page 112]

that will yield benefits for insurance companies and policyholders alike, it also can be used to violate policyholders' rights. We have seen insurance companies do this before and it already appears to be happening again. The time for vigilance is now.

Recent use of new algorithms to deny medical claims by seriously ill and elderly patients underscores the need to carefully examine what is at stake for policyholders. Insurance companies must uphold their duty of good faith and fair dealing. Insurance companies must not use AI and algorithms to prioritize operational efficiency at policyholders' expense.

Insurance (and banking) regulators already are highlighting problems posed by the "lack of explainability" in AI tools under development. ¹ Watchdogs should take stock of past misdeeds and ensure that they are not repeated to an even more harmful extent by AI.

Lesson from the Past

Twenty-five years ago, an emerging technology called Colossus was used as a claims-adjusting tool by many insurance companies, including Allstate. This database system evaluated personal injury claims and generated a range of potential settlement amounts based on comparable past injuries. Colossus allowed claims adjusters to use their discretion when using the software, and some insurance companies leveraged the software to fine-tune results in order to increase profits at the expense of injured policyholders.

An in-depth investigation by policyholder attorney David Berardinelli revealed that Allstate tweaked the software to produce lowball claim estimates and incentivized claims adjusters to depend exclusively on the estimates recommended by Colossus. Allstate was sued for its claims-settling protocols and ended up paying millions in damages and penalties for its bad faith practices, thanks to Berardinelli's hard work and persistence in forcing Allstate to turn over incriminating documents.

In the mid-1990s, Allstate hired McKinsey & Company to redesign Allstate's claims system in a project termed the Claims Core Process Redesign (CCPR). ² This revamping aimed to bolster Allstate's profits through an aggressive reduction of claim payouts.

[Page 113]

The plan for the CCPR was summarized in a series of more than 12,000 PowerPoint slides used by McKinsey to explain the plan to Allstate personnel and executives.

Berardinelli sought to uncover the McKinsey slides, but Allstate fought the disclosure tooth and nail. Berardinelli, however, eventually convinced a judge to compel Allstate to turn over the slides and, eventually, to make the documents public. Allstate still resisted these court orders and was held in contempt and fined for its willful disobedience in refusing to turn over the slides.

Once the McKinsey slides went public, it was glaringly apparent as to why Allstate was determined to keep them hidden. Part of McKinsey's strategy—known as "Good Hands or Boxing Gloves"—created a dual approach to handling claims: offering a seemingly quick settlement option for compliant policyholders...