Insurance matters can be frustrating at times. For example, the Twitter user Reinhard Remford reported on the online portal Twitterperlen that he wanted to report a claim and to do so, registered on the online portal of his insurer. There he only found a telephone number for reporting the damage. When he called, a tape announcement sounded there – with the correct Internet address for the damage report.
The case shows that some insurers still have a lot to catch up on in order to keep up with customer requirements. Insurers have high hopes for artificial intelligence (AI). The software is supposed to sift through large amounts of data, learn from it and make decisions independently. This should relieve the employees of standard inquiries and offer customers a better and faster service – be it for a change of address, which the chatbot receives and notes in the system in real time, or a request for residential building insurance, for which only a few details are required because the AI can use the address of the property to be insured to identify many things itself. Benedikt Kalteier, CEO at Generali, is primarily concerned with understanding customers better with the help of AI. “Insurers have still not understood what customers want,” he said at a SZ conference. Technology could provide information on product design and pricing.
Provincial board member Patric Fedlmeier sees the greatest benefit of AI in claims processing. Instead of clicking through many portals, the customer simply sends a picture of the damaged object, which the AI analyzes. This is not only more convenient for the customer, it can also pay off for the insurer. Depending on the line of business, 50 to 70 percent of the costs are claims costs. “If you could get a percentage point out of that, you would have a real leverage,” says Fedlmeier. The providers also have high hopes for combating fraud with AI.
It is a big task to use the available data
Despite its many advantages, AI is not yet widely used in the insurance industry. “Not everyone has yet understood what power is behind it and what framework conditions are necessary for the use of AI,” said Fedlmeier. Insurers have the large amounts of data required by technology in-house, but usually not in a form in which AI could access them. It is a big task to make this data usable. “We have a thick board to drill,” he explained.
Michael Zimmer, Chief Data Officer at Zurich Group Germany, believes that insurers are already using a lot of AI, but this is not obvious at first glance. “We do AI,” he said. “But we have them in too many different systems.” It is important to bring the AI applications together – and also to talk more about them. “We are bad at selling our innovations positively,” believes Zimmer. It is important to show what the insurers are already doing – also to attract the right employees. Because for aspiring data analysts, an insurer is not exactly the first choice as an employer.
One thing is certain: insurers cannot completely ignore the new technical possibilities. “This separates the wheat from the chaff,” said Daniel Besendorfer, CEO of Allianz subsidiary Allianz Technology. Those who get involved with AI & Co. will be successful and don’t have to be afraid of internet giants like Amazon. “If you refuse, you will get into trouble,” he believes.