Insurance is part of everyday life. The concept provides not only peace of mind but protection and mitigation to the risks of everyday living. Take into consideration car, life, health, and even natural disaster insurance and suddenly the coverage and amount of claims going through everyday can be daunting. There is also the risk of fraud, changing premiums, and the expectations of life long customers and it becomes clear the difficulty insurance providers are under. Luckily artificial intelligence is making not only the lives of insurance companies easier but their customers as well.
When looking at how artificial intelligence is aiding insurance, it’s best to look at how claims are being affected. A claim is any formal request made by the policy holder to an insurance company for the coverage or compensation for a covered policy. The insurance company then validates the claim and once it is approved the policyholder is paid. So how is AI making this better? That can be answered by looking at how AI is increasing the speed at which claims are assessed and solving both new and old problems.
One of the easiest ways to send in a claim is by email, the only real issue is that a person has to answer each and every email. Depending on the claim and the person answering them this can start to add up, at least it used to until the company Cognizant designed an AI to fix this problem.
Cognizant explained that their AI is designed to, “extract policy, claim, and agent-related information automatically from unstructured incoming emails and attachments. The solution integrates our client’s mail exchange server with a cognitive engine on Microsoft Azure.” It was stated that the AI is capable of this due to it’s, “AI-driven self-learning pattern recognition and keyword text recognition to extract data automatically and then respond without manual intervention.” Cognizant went on to say that this was done in order to enable the resigning of human labour to more valuable tasks.
Basically, this AI processes a large number of emails and sorts them out while obtaining the most important details from what it's learned from past experiences. It then approves or denies claims in a timely manner. If there is an anomaly or very unique case it relays the claims to an expert to make the final decision. According to Cognizant, this adds up to 60% cost savings, a 50% increase in ability to handle inquiries and six times the increase in the speed of customer service.
Not only is AI helping to solve an increase in email claims but also helps tackle fraudulent claims as well. Fraud adds up, according to the Insurance Institute of Canada automotive fraudulent claims cost taxpayers 1.6 billion annually. To understand artificial intelligence's ability to stop fraud it's best to look at a 2011 fraud case out of the U.K. where criminals faked 120 car accidents to claim over £1 million in false claims. This was a major problem, but Luca Lanzoni, Chief Information Officer at HDI Assicurazioni and a powerful AI had a solution.
When Lanzoni took this issue head on he used an advanced AI system. It was explained that the AI “was able to cross-reference individuals and companies through internal and external fraud records, news sites, credit, and law enforcement databases and sift through claims histories in seconds.” It then found patterns through accident claims and news reports that showed the criminals were crashing similar old vehicles into brand new BMWs in the same ways over and over again. These details were then given to agents to make a final decision, which led to the criminals' arrests.
This shows that AI can help insurance agents properly utilize their time, these programs pick up on patterns across large bodies of data at a speed that might not be possible by humans because of the size of it all. An expert still has to go over the patterns to determine if they are suspicious and worth looking into or just a one-time thing. This AI won’t replace employees, they will however help protect the company and the law.
Cognizant and Sandilands aren't the only ones to point out the benefit of AI in claims. A New York based insurance company Lemonade has been using AI to process claims faster than ever before. The company has been quick to point out an example of theirs where “the company paid a claim for a stolen $979 Canada Goose jacket in just three seconds.” This may not seem like a big deal but the idea that a larger sum of money for a claim was that easy to process and done at that speed can make a huge difference in someone's life, not to mention bring in repeat business.
In a quote from the Financial Times, the CEO of Lemonade Daniel Schreiber has stated that “a bot handling claims in second’s delights customers and crushes costs.” He has also explained that 11%-13% of the premiums paid by customers go to the bureaucracy of handling claims, and that AI cuts down on those costs. This means that people are excited to try out an AI and that the work they do is invaluable to insurance companies that want to save both money and time. In a world market where every penny counts an AI could mean all the difference.
What It All Means
AI is advancing in almost every field and insurance isn’t any different. What is different is the added benefit of AI in insurance claims. Whereas most companies use AI to make things more efficie