It is expected that by 2026, over 100 million humans will engage with gen AI to contribute to enterprise work.1 One industry that will see huge gains in efficiency, cost savings and improved customer satisfaction through the adoption of AI is the insurance industry.
By 2027, it’s estimated that 25.5 billion interactions will be resolved using AI automation, compared to 2.7 billion in 2022.2 So, what will gen AI do for the world of claims and what does this mean for insurers?
Claims notifications whether they are emails, online claim forms or even a chatbot can be read and created by smart robotics, and gen AI can take this one step further by ascertaining context from the notification to complete any missing details.
As with human claims handlers, AI can infer from context while working within a set of rules, what the presented information was trying to convey. For example, if in a ‘what happened’ box the claim referred to flooding from a damaged washing machine but in the damage section of the claim form the washing machine was omitted, AI can add that information from the context and supplement this with additional external data sources.
The growing intelligence and increased adoption of gen AI has the potential to revolutionise claims administration over the next few years. AI and machine learning have the power to automate many tedious and labour-intensive tasks in the claims process. In doing so, gen AI can augment human claims handlers to:
Improve efficiencies for faster claims processing
Reduce costs associated with claims handling
Mitigate human errors and inaccuracies
Enhance the customer experience and customer satisfaction
Detect and prevent fraudulent claims
Gen AI’s role in policy-based insurance claims
As a human claims handler would check facts such as date, cause of loss or extent of damage; gen AI is equally able to cross-check these facts against its knowledge base and integrated systems. The most likely integrations include:
Street views and maps for motor insurance claims
Medical journals for life insurance claims
Weather data and news articles for home insurance claims
Once a claim has been logged, AI can check the policy wording for most policy led insurance types such as home, motor, life and pet insurance. AI will then confirm whether cover should be applied and knowing the extent of cover, can issue a communication (via email, SMS or other means) to the claimant with a request for further information or a letter to confirm if the claim is covered or not.
In instances where claims are covered, AI can then calculate the costs of replacement, deduct any applicable excesses, make an offer to the claimant, and possibly even issue a cheque if the claimant had previously indicated that they have replaced the goods and are just looking for reimbursement.
Over time, AI will be able to automate the claims processing workflow, learn from old claims and identify patterns in new claims. There isn’t a lot a smart AI with claims training and experience won’t be able to complete. This will significantly reduce the burden on human claims handlers, freeing up their time to focus on other core areas of the business and accelerate claims processing time to improve the customer experience.
Gen AI’s role in liability-based insurance claims
For the most part notification of a liability claim is much like any other. The area in which liability-based claims differ from policy-based claims is the investigation process, which is not as document heavy and relies much more on handler knowledge and instincts.
So, will an AI be able to handle a liability claim for let us say an employer’s liability claims for an injury which occurred at work while the employee was on break? While your first thought would be no, this claim requires handlers’ instincts to understand the proximate cause of the accident, the legal precedents that come into place as well as all the various laws which sit around employment and employee safety.
Liability-based insurance claims are typically multifaceted, and as such handler knowledge is required to navigate complex claims such as the above. However, this would not restrict the use of AI.
AI’s have recently passed many academic examinations including degree level legal qualifications as well as the Bar examinations.3 This proves the ability of AI to understand and apply case law and precedents to legal arguments, which in turn removes much of the argument for human handlers on liability claims.
Combine this with the human-like ability to check images, street views and maps to understand the circumstances of the incident, AI can not only competently handle liability claims but also identify disparities in the allegations to highlight potential fraud. Once liability decisions have been made the rest of the handling process would be extremely simple for AI to process.
[1-2] Gartner Market Impact: AI and Automation disrupt traditional customer management BPO solutions, January 2023
