Could AI address ‘pain points’ in claims service for customers?

Artificial intelligence’s (AI) time in the public and professional spotlight is not a new phenomenon.

AI development stretches back to 1955, when the term “artificial intelligence” was first coined in a proposal for a “two month, 10 man study of artificial intelligence” conducted by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon.

The workshop, which took place a year later in July 1956, has been generally considered as the official birthdate of the new AI field.

When, in 1965, Herbert Simon predicted that “machines will be capable, within twenty years, of doing any work a man can do,” many scoffed at the idea. 

However, the shockwaves of the pandemic proved Simon to be somewhat right.

As a result of the pandemic, AI has surged into centre stage and onto the lips of many, captivating a wide range of industries.

Many will have encountered the increasingly common question arising in business, “have you integrated AI into your operations?”.

However, business leaders have frequently found themselves pondering whether AI is a revolutionary force or merely based on exaggerated expectations.

The insurance claims sector has not been immune to this wave, with industry leaders actively investigating how AI can elevate their companies.

Dewald Botha, director of AI at ExploreAI, attributed the seismic shift of the last 12 months to advances in technology and the introduction of OpenAI’s ChatGPT.

During Insurance Times’ AI: Revolution or Hype webinar, chaired by acting editor Yiannis Kotoulas earlier this month (5 October 2023), Botha said advancements in generative AI, specifically around algorithms and models, had taken massive steps forward.

Botha explained that algorithmic advancements occurred in 2017 and that the deep learning framework was “instrumental for setting generative AI up for success.”

He explained that this came in “combination” with data generation, accessibility and the surge in computing power, which, when combined, enabled the training of “colossal” models on extensive datasets.

Botha added that the public release of ChatGPT in November 2022 was the first time the public had access to AI through “such a user-friendly interface.”

This experience created a “marketing boom,” he said, drawing the interest of the insurance sector and wider industry.

Implementation

The explosion in the awareness of AI, its accessibility and its potential use cases captured imaginations in boardrooms and spurred them to explore whether it could improve their businesses.

As David Carmalt, managing director of ExploreAI, noted: “The explosion into public consciousness has bled into the c-suite.”

He explained that corporate leaders were now beginning to ask questions about AI’s relevance to their businesses and distinguish between hype and reality.

“From a corporate perspective, this is the first time we’ve seen a major move at a c-suite level to engage in these discussions and at such a wide scale.

“Senior leaders within financial services are beginning to understand the impact that data science can have on the bottom line – whether driven by improvements in customer engagement, process automation, risk pricing, reserving, fraud detection or other areas.”

Apoorv Kashyap, a consultant at EY, pointed out that pre-trained models and accessible AI technologies had opened doors for real-world applications in the insurance industry.

He noted that “AI could change businesses,” but decisions on how and where to implement the technology should be grounded in tangible needs and expected results.

Kashyap added: “My biggest advice is to start with the lowest common denominator – what are your biggest pain points? From this, you usually get the biggest return.”

Pain points

The insurance industry has been particularly focused on employing AI to address pain points, with claims processing standing out as a prime example.

Earlier this year (14 September 2023), the Financial Ombudsman Service (FOS) revealed that complaints over building, car and motorcycle insurance had hit a five-year high in the first three months of the current financial year.

Some 3,869 complaints were made about car or motorcycle insurance, while there were a total of 1,776 building insurance complaints recorded during the period.

By comparison, the FOS recorded 2,626 motor and 1,275 building complaints between April and June in 2019.

In response to the heightened complaints numbers, the regulator said that it expected insurers to treat their customers fairly and promptly – and that it was “unacceptable” that complaints were driven due to a delay in claims being paid out.

Meanwhile, according to Catherine Carey, head of marketing at Consumer Intelligence, over 60% of claimants cited speed and efficiency as the element of the claims experience needing the most improvement.

In this context, Kashyap said this was AI’s “sweet spot”, particularly in claims processing, where there is a substantial volume of data processing and information exchange.

This is the ideal scenario for generative AI, as it’s precisely what it has been pretrained for, and many in the insurance industry have been using it for this exact reason, Kashyap added.

Insurers have recognised this potential, with Zurich experimenting with AI for tasks like data extraction for claims and modelling earlier this year (27 March 2023).

Insurtech Lemonade notably managed to pay out a claim in the UK without any paperwork in just two seconds with the help of its AI chatbot earlier this year.

Meanwhile, Minster Law introduced a new platform to streamline its claims registration process, featuring an intelligent injury matrix for automated claims track allocation and tailored customer inquiries.

Paul Taylor, chief operating officer at Minster Law, highlighted that the new platform had been designed to “respond to the needs of today’s time-pressed customer.”

More recently, Sedgwick said it had been using artificial intelligence (AI) to enhance the claims experience for customers.

Richard Sheridan, strategy development director at Sedgwick, told Insurance Times that AI could be leveraged to automate aspects of claims information collection.

For example, he noted that data analytics, combined with behavioural science, could help the company “better understand and predict a customer’s needs”, allowing it to design systems that “anticipate and proactively respond”, rather than simply reacting to a customer request.

Sheridan said: “We can use AI to automate parts of claims information gathering where we already have the data or it’s available through various external sources. This speeds up the process and avoids customers having to provide us with details we can obtain automatically.

“Sophisticated omnichannel communication platforms also ensure that customers can seamlessly move between online, phone or self-serve channels, with a full view of all interactions with claims handlers.”

“AI is brilliant at processing repetitive tasks. It can interpret rules, share data across different functions and carry out a wide range of basic administration tasks.”