Artificial Intelligence is Disrupting the Insurance Industry

What exactly is artificial intelligence?

According to IBM, who has been doing research on the topic since the 1950s, artificial intelligence (AI) is simply “anything that makes machines act more intelligently”.­ This can be broken down further into two groups: applied or general. Applied AI is capable of doing only those tasks it has been designed for, such as driving/operating a car, or trading stocks. This is the most common type of AI, and has had a wide range of success in many applications. On the other hand, general AI is far more advanced and theoretical. Instead of seeing specialization in one task, “a machine [would be] able to perform any task a human can”. It would, in essence, be able to learn anything and apply it to any situation, to think and reason just like a human. This type of artificial intelligence, though generating much attention in the recent years, is still far from a reality.

Hand-in-hand with the concept of AI is the area of machine learning (ML). This is “a branch of artificial intelligence based on the idea that systems can learn from data, identify patterns and make decisions with minimal human intervention”. This technology is widely used today, mainly by industries who amass a large amount of data (e.g. financial services, government, health care, etc.). It is what gives a machine its ability to learn from the past, and make predictions/decisions for the future. The more it encounters, the smarter it becomes.

The financial services industry has been using artificial intelligence and machine learning for years. Referred to as FinTech, firms take advantage of AI’s power by providing sophisticated virtual customer service assistants, analyzing legal documents, and even making strategic trades on the stock market. Large firms such as JP Morgan Chase have adopted artificial intelligence to help analyze important documents, which in turn has reduced their time spent from 360,000 hours annually down to a matter of seconds.  With AI’s ability to streamline back-end operations and improve efficiency all around, one might wonder why other industries have been so late to the game. Insurance firms, for example, are only just beginning to experiment with the use of AI. Considering its resemblance to the financial industry, it seems odd that insurance is still lagging so far behind. There may be, however, two reasons why FinTech got a head start. For one, insurance is a very passive product. Whereas roughly 70 percent of insurance firms only hear from their customers once a year, financial institutions communicate with their clients almost 200 times annually. This frequent customer interaction, coupled with the heavy regulation following the economic crisis of 2008, created a need for change. It also created the perfect opportunity for disruption from startups and innovators. “From non-bank lending, because banks could no longer provide enough capital, to consumer friendly apps and efficient payment solutions” the introduction of FinTech was unsurprising. Given that the field has steadily grown over the years, and amassed $16.6B in investments for firms last year, AI’s possibilities should not be ignored. The question is: can other industries catch up?

For the insurance industry, AI and machine learning have thus far been used in pricing, handling claims and detecting fraud, though firms are only now learning of its endless possibilities. Lemonade, a new property insurance company out of New York, is pioneering the way for other carriers to automate their processes and implement artificial intelligence. They enlist the help of chatbots, which are computer programs able to analyze language and mimic conversations to interact with humans. The bots are able to speak to multiple customers at once and are available anytime of the day. This is a major advantage for businesses that rely on fast, effective customer service.

Lemonade’s chatbot, Maya, “sells inexpensive homeowners’ and renters’ insurance, and their claims bot, AI Jim…recently settled a claim in three seconds”. Maya is able to communicate directly with customers and help them navigate through the confusion of applying for coverage. Their website claims she will craft the perfect insurance for you, without the need for customer service representatives and underwriters. The appeal of having a computer execute such tasks is that it reduces the time spent, the hassles and the costs. It makes the processes far more seamless, instantaneous and trustworthy. Even though Lemonade is targeting a niche group (millennials), other insurance companies should not turn a blind eye; they could learn a lot from the company’s operations and values. “Lemonade is fast and transparent rather than slow and opaque”. Many people will be drawn to this refreshing view in a notoriously mundane industry.


With the ever-increasing use of social media and “smart” gadgets, AI machines now have access to a wealth of data. This is especially useful when analyzing a client’s risk and setting an appropriate premium. Underwriting, a somewhat lengthy and intricate process, has been presented with the opportunity for automation. A bot is able to “scan a customer’s social profile to gather information and find trends and patterns”. This ability to analyze social media posts and determine a person’s risk within seconds puts AI’s capabilities far beyond humans’, and at a far lower cost. When considering the vast amount of data generated from the Internet of Things (IoT), the accuracy and knowledge of these bots will be unparalleled. AI also provides the “mechanics to capture ‘tribal knowledge’, thereby providing a uniform assessment metric across the entire underwriting process”. Tribal knowledge is defined as information that is known only to the insiders of a particular group or organization, and is not common knowledge to outsiders. In the world of underwriting, this tribal knowledge could consist of emails, internal reports, presentations and evaluations, all of which can help better assess a group’s risk.

One would think gathering personal data from people would be met with overwhelming opposition, but surprisingly “62 percent of younger groups said they’d be happy for insurers to use third-party data…to lower prices”. While it is exciting that companies may no longer base their premiums on generalized assumptions, and look to individualized data instead, a concern for privacy is presented. Allowing limitless access to your personal data is both risky and invasive. What happens to our freedom of speech if we can no longer share photos from a night out, or post our thoughts online without the threat of increased insurance premiums? Additionally, think of the impact a data breach would have if you allow all of your personal information to be in the hands of one company. To make matters worse, there is no guarantee these companies would not sell your information to third parties, thereby increasing the risk of a data breach exponentially. Consumers need to weigh the pros and cons before allowing insurance companies access to such personal information. Since when does affordability outweigh privacy?

Claim Handling

Claims processing can be “a monotonous task susceptible to errors stemming from uniquely human factors”. Not only does AI reduce the amount of time spent on claims, it also reduces the probability of error. Imagine having a machine analyze a photo of damage and estimate repair costs within seconds. The entire burden of handling the claim would fall off of the insurer and customer. AI will eliminate the need for having multiple people work on the same claim, thus decreasing administrative costs and the frequency of errors. Kristof Terryn, COO of Zurich, has instituted a project to automate their claims processes. He declared that it “will trim $5 million from expenses…for the 39,000 hours of claims handler time now being done by computer”. Though this is an impressive reduction in costs, most of the value stems from increasing the accuracy of claims and virtually eliminating errors during the process. Other companies that have implemented automation “of their claims process have seen a significant reduction in processing times and [an increase in] quality”. The use of chatbots also reduces the need for interaction with customers during the process. Any questions customers may have could be directed to those bots, therefore allowing workers to focus on more important tasks. Thanks to artificial intelligence, insurance companies’ resources can be better allocated; they will no longer require the manpower needed in decades past.

Fraud Detection

With the implementation of AI and machine learning comes the ability to analyze data better and faster than any human ever could. These machines are able to identify patterns within a claim, look to historical data and “help to recognize fraudulent claims in the process”. This should be of utmost importance to not only insurance carriers, but to the insureds as well. According to the FBI, it is estimated that insurance fraud costs the United States $40 billion annually. This cost is passed along to the insureds in the form of increased premiums, roughly “$400 to $700 per year”. A startup firm in France called Shift Technology is using AI in their fraud prevention services. They have “already processed over 77 million claims… [and] have reached a 75 percent accuracy rate for detecting fraudulent insurance claims”.1Even IBM offers AI services to fight financial crimes.  It is evident that this application of AI will grow exponentially in the future and drastically help cut the cost of fraud.

All things considered, artificial intelligence is set to flip the insurance industry upside down. Automating monotonous processes, reducing fraud and increasing accuracy are major advantages of using the technology. Insurance, which as an industry is notorious for its antiquated processes and values, is being disrupted by this emergence of artificial intelligence and machine learning. Disruption happens when an existing market, industry, or technology is displaced and replaced with something new and more efficient/worthwhile. Accenture claims that “75 percent of insurance executives believe AI will provide significant industry changes in the next three years”. Despite the fact some insurers have started to embrace the change, many others are unaware of the impact it is having already. While there are many kinks left to be ironed out, the reality is that companies have two options- catch up or be left behind.


About the Author

Ashley Evans: Analytics intern at J.W. Terrill, and senior at Saint Louis University pursuing a bachelor’s degree in Finance; Dean’s List recipient for two consecutive semesters. Interested in continuing her education with a Master’s in Applied Financial Economics.


  1. “Artificial Intelligence.” IBM. July 25, 2016. Accessed July 18, 2018.
  2. Marr, Bernard. “What Is The Difference Between Artificial Intelligence And Machine Learning?” Forbes. September 15, 2017. Accessed July 18, 2018.
  3. “What Is AI?” IT PRO. June 11, 1970. Accessed July 18, 2018.
  4. “Machine Learning: What It Is and Why It Matters.” SAS. Accessed July 18, 2018.
  5. Alton, Liz. “How Financial Services Use AI To Serve Customer Needs.” Forbes, Forbes Magazine, 8 Sept. 2017,
  6. Kurani, Ravi. “InsurTech Is the New FinTech – or Is It? – Earlybird’s View – Medium.”Medium, Augmenting Humanity, 23 Aug. 2016,
  7. “The Global Fintech Report Q4 2017.” Corporate Innovation Trends, 2018, Ads&utm_medium=FintechReport&msclkid=4a5f3b8d528b149feb2660d5b3c03485.
  8. Team, Writer Profile Of “How Artificial Intelligence Is Changing the Insurance Business.” Medium. February 14, 2017. Accessed July 18, 2018.
  9. Sun, Alex. “How Chatbots Can Settle an Insurance Claim in 3 Seconds.” VentureBeat,VentureBeat, 28 May 2017,  settle-an-insurance-claim-in-3-seconds/.
  10. “Lemonade Renters & Home Insurance | Protect The Stuff You Love.” Lemonade. Accessed July 18, 2018.
  11. Fromm, Jeff. “How Startup Lemonade Is Redefining Insurance For Millennials.” Forbes. July 12, 2017. Accessed July 18, 2018.
  12. Moodie, Alison. “How Artificial Intelligence Could Help Make the Insurance Industry Trustworthy.” The Guardian. January 28, 2017. Accessed July 18, 2018.
  13. Morgan, Blake. “How Artificial Intelligence Will Impact The Insurance Industry.” Forbes. July 25, 2017. Accessed July 18, 2018.          intelligence-will-impact-the-insurance-industry/#1daeb20a6531.
  14. Kumar, Krishna. “AI’s Huge Potential for Underwriting.” Insurance Thought Leadership. December 17, 2015. Accessed July 18, 2018.
  15. “What Is Tribal Knowledge? Definition and Meaning.”   Accessed July 18, 2018.
  16. “Young Consumers Willing to Let Insurers Spy on Digital Data – If It Cuts Premiums.”     Insurance Journal. June 21, 2018. Accessed July 18, 2018.
  17. “Insurance Fraud Prevention Gets Help From Artificial Intelligence.” Samsung Business Insights. January 19, 2018. Accessed July 18, 2018.
  18. “AI and Insurance: Are Claims Jobs in Danger?” Carrier Management. March 14, 2017. Accessed July 18, 2018.
  19. “Insurance Fraud.” FBI. March 17, 2010. Accessed July 18, 2018.
  20. “RegTech and Cognitive Risk & Compliance | IBM.” IBM Cognitive Advantage Reports.             Accessed July 18, 2018.   markets/risk-compliance.
  21. Howard, Caroline. “Disruption Vs. Innovation: What’s The Difference?” Forbes. June 20, 2016. Accessed July 18, 2018.
Linkedin Facebook Twitter Email

No comments yet.

Leave a Reply

Time limit is exhausted. Please reload the CAPTCHA.