The ethical implications of automating high level claims processing

22 March 2023 / 5 min read

Claims differ in severity and human impact – from lost equipment covered under content insurance through to car accidents – and customers expect different levels of sensitivity from their insurer, so the goal for insurers must be to ensure any AI or automation introduced into this process complements a human-driven, personalised Customer Experience (CX).

ethical-implications

The level of claims overseen by human operatives is predicted to drop by up to 90%

Automated systems can categorise claims, gather information, produce expert decisions and deliver the next best action. For example, insurers can process claims for water damage to home interiors utilising AI throughout the claims journey – from smartphone app usage to capture images for initial response, to AI driven image analysis for decisions on salvageability and associated costs.

But with the level of claims overseen by human operatives predicted to drop by up to 90% – insurers face a difficult ethical dilemma when using AI.

Digital Processing – the defining insurance market trend of the post-pandemic era

Digital processing played a significant role in helping insurers maintain operations during the pandemic slowdown, as it allowed a dual human-systems approach to mitigate against the lack of physical, in-person CX. In fact, 85% of insurance CEOs also credit the pandemic with growing the necessary technologies for digitising operations and creating new efficiencies.

In the post-pandemic era, this has allowed businesses to redirect human resources to areas of business they were most required, and activities most beneficial to their development.

Employees happy with hybrid-systems approaches

Additionally, amid the hybrid working revolution, digital processing has increased retention rates and job satisfaction by allowing employees freedom to work from home when necessary. Market leading insurers now offer both virtual and physical underwriting rooms, meeting the 26% rise in split home-office working compared to pre-pandemic levels with digital processing. Such operational models for claims processing can attract new talent and secure the long-term future of business.

But some policyholders are less eager for the AI revolution

There are huge benefits for insurers willing to embrace AI to improve CX, and although 73% of consumers agree their purchasing decisions are influenced by CX quality, 42% are willing to pay a premium for friendly, ergonomic CX. The customer is king in the insurance technology space, and in studies from Oxford University on the proliferation of AI, 34% of respondents reflected extremely negatively. Additionally, the FCA’s Consumer Duty standard has recently outlined legislation to support consumer power and enhance customer protection across CX – so sensitively integrating automation into CX has never been more important.

Insurers should proceed carefully with AI in claims processing. Ensuring the policyholder will be comfortable with the technology being used is just as important as appeasing the policyholder with a fast and effective digital claims platform!

Identify vulnerability – the ethics of handling claims

Making a claim often renders a customer as vulnerable. The person may be suffering shock following a serious car accident or emotionally affected by serious property damage. The application of AI for claims processing is not appropriate for every customer – AI and automation must be tailored to the context and the customer, and the insurer has to ensure the correct physical or digital CX is in place.

If the insurer can divert these high-level claims to a human claims handler to observe the source of vulnerability at its root, they are able to support individuals appropriately. Technology is critical to growth but using it tentatively alongside a traditionally human-driven claims handling system, must be the standard with the practices of most major insurers.

AI is transitioning

A case can be made for innovatively using empathy replication in AI systems to sympathetically extend the use of digital processing to full automation. With technology for emotional synthesis emerging, AI is transitioning from a system complementing human emotions, to a system capable of replicating human emotions – even anticipating interactions accordingly.

Whichever operational response insurers choose, the creation of positive end-customer experiences should be central. Insurers should acknowledge the customer sensitivity issues when using AI, and that means having the infrastructure to provide the high-quality CX critical to brand retention and growth.

The future of the use of data in claims handling

The general industry assumption is that characteristics such as gender and age should be kept away from models for AI, but this can be difficult in scenarios reliant on personal data – such as the property information required for remote damage assessments. Some consumers may be reluctant to offer AI sensitive data, and the process must therefore be visible, understandable, and transparent to the policyholder.

Since 2018, insurers have been legislatively bound by law to use customers data for only explicit purposes, with additional tight controls over claims critical factors such as health records and criminal convictions representing significant obstacles for insurers in the AI process.

However, as outlined in the 2022/23 revisions to the Data Protection Bill, the House of Commons is committed to simplifying online identity proofing to allow the digital identity marketplace to thrive – presenting opportunities for insurers to develop AI-informed customer profiling and speed up the claims process once the bill is passed. With 69% of policyholders willing to share sensitive health and exercise data to insurers if it reduces their rates, the marketplace is likely to boom in the coming years.

Balance AI and human service to digitally engage customers into the future

Critically, insurers need to find the right balance between AI and human customer support. Customer expectations are rising, insurers need to adapt by pre-empting industry change and proactively preparing systems to better and seamlessly serve the policyholder. Most policyholders are funnelled towards digital platforms, but this still needs the required levels of empathy and personalisation.

Verisk is at the forefront of bringing automation to insurance claims using AI and predictive analytics, but with a core focus on keeping the personalised customer service at the heart of these processes. Click here to see how to bring speed and precision to your claim decisions while maintaining the human touch.