How ar­ti­fi­cial in­tel­li­gence is trans­form­ing in­surance

AI is giv­ing in­surance new ways to enage with their cus­tomers, process claims and ad­just risk.

Digital Insurance - - VOICES - BY JOE MCCKENDRICK

The in­surance in­dus­try seems like a nat­u­ral set­ting for ar­ti­fi­cial in­tel­li­gence to take root. So­lu­tions pro­vid­ing fraud anal­y­sis against claims data has been in place for some time at many com­pa­nies, en­abling rapid set­tle­ments with a high de­gree of as­sur­ance that fraud­u­lent claims can be spot­ted be­fore checks are cut and de­liv­ered. There’s also noth­ing new about au­to­mated underwriting at the front end of the pol­icy life­cy­cle, as­sur­ing that most rou­tine ap­pli­ca­tions are pro­cessed and priced in the most ex­pe­di­ent man­ner.

Yet, it ap­pears the in­dus­try is set to move to an even more com­pre­hen­sive gen­er­a­tion of AI, one that may trans­form in­surance in untold ways. This is set to hap­pen very quickly — not in some far-off fu­ture. That’s the gist of a re­cent Ac­cen­ture study, which finds that in­surance ex­ec­u­tives be­lieve that AI will sig­nif­i­cantly trans­form their in­dus­try in just a few three years.

Ac­cord­ing to the re­port, three-quar­ters (75%) of in­surance ex­ec­u­tives be­lieve that AI will ei­ther sig­nif­i­cantly al­ter or com­pletely trans­form the over­all in­surance in­dus­try in the next three years. Close to one-third (32%) be­lieve that their own com­pany will be “com­pletely trans­formed” by AI within that time­frame. The re­port is based on the in­sights of a tech­nol­ogy ad­vi­sory board, in­ter­views with in­dus­try tech­nol­o­gists and a sur­vey of more than 550 in­surance ex­ec­u­tives across 31 coun­tries.

So, with one-third stat­ing their com­pa­nies will be com­pletely al­tered by AI, where and how is this set to hap­pen? What will it look like? The au­thors state that AI will be­gin to shine as they be­gin to in­vest in AI to em­power agents, bro­kers and em­ploy­ees to en­hance the cus­tomer ex­pe­ri­ence with au­to­mated per­son­al­ized ser­vices, faster claims han­dling and in­di­vid­ual risk-based underwriting pro­cesses.

What does not seem likely is a rise of ro­bot-run in­surance com­pa­nies, with just a few hu­man op­er­a­tors. The re­port’s au­thors are very care­ful to state that AI will “em­power,” ver­sus “re­place” hu­mans. Some in­dus­try an­a­lysts have spec­u­lated that many agents, bro­kers and em­ploy­ees will find them­selves pushed aside by au­to­mated and highly in­ter­ac­tive cus­tomer ser­vice tech­nol­ogy. If any­thing, Ac­cen­ture states, the rise of AI is cre­at­ing new op­por­tu­ni­ties for in­surance agents, bro­kers and em­ploy­ees “to de­liver a bet­ter cus­tomer ex­pe­ri­ence, with the tech­nol­ogy en­hanc­ing the way sales and ser­vices are ex­e­cuted, fa­cil­i­tat­ing faster claims pro­cess­ing, and en­abling more-ac­cu­rate, in­di­vid­ual risk-based underwriting pro­cesses.”

Where AI will be mak­ing its mark is in the way users both in­side and out­side or­ga­ni­za­tions in­ter­act with tech­nol­ogy. “AI is the new UI” (user interface), they state, adding that “we are see­ing the rise of the ‘in­tel­li­gent in­surer.’” AI is giv­ing rise to tech­nol­ogy that “is in­creas­ingly in­ter­ac­tive, as touch dis­plays, mixed re­al­ity, and nat­u­ral lan­guage pro­cess­ing make it feel more hu­man. Ad­vanced tech­nol­ogy is now ca­pa­ble of learn­ing, with con­tex­tual anal­y­sis, im­age recog­ni­tion and deep learn­ing al­go­rithms that make it seem to think more like us.”

Al­ready, AI is be­ing em­ployed “to stream­line claims, an­swer ba­sic cus­tomer queries and, in­creas­ingly, to of­fer straight­for­ward ad­vice about com­plex prod­ucts to cus­tomers in a cod­i­fied and con­sis­tent man­ner.” The un­der­pin­ning tech­nol­ogy that makes AI pos­si­ble in­cludes au­to­mated data cap­ture and recog­ni­tion, ro­botic process au­to­ma­tion and cog­ni­tive robotics.

AI is al­ready in use in many as­pects of the in­surance in­dus­try. Claims ad­justers are us­ing it to sim­plify the triage process. Ini­tia­tives like MassMutual’s Haven Life use it to sup­port fast, sim­ple life in­surance sales and pol­icy writ­ing. IBM Wat­son sup­ports Fukoku Mu­tual Life In­surance’s call cen­ter in Ja­pan, rout­ing calls based on nat­u­ral lan­guage pro­cess­ing

And Ge­ico has a vir­tual as­sis­tant — “Kate” — on its mo­bile app. “She” un­der­stands nat­u­ral lan­guage and can an­swer ba­sic ques­tions like: “What is the cur­rent balance on my auto in­surance pol­icy?” or “When is my next pay­ment due?”

Mov­ing to AI is not without its chal­lenges, and the Ac­cen­ture team ex­plored th­ese as well. The re­port found that in­sur­ers face chal­lenges in­te­grat­ing AI into their ex­ist­ing tech­nol­ogy, cit­ing is­sues such as data qual­ity, pri­vacy and in­fra­struc­ture compatibility.

A press­ing chal­lenge not ad­dressed in the Ac­cen­ture re­port is the avail­abil­ity of skilled per­son­nel who can put AI in place and pro­vide on­go­ing over­sight and en­hance­ment. The in­surance sec­tor will be com­pet­ing with tech com­pa­nies, uni­ver­si­ties and other seg­ments for what is still rare and valu­able tal­ent to pur­sue AI ini­tia­tives, as well as re­lated dig­i­tal ef­forts.

Joe McKen­drick is an au­thor and in­de­pen­dent an­a­lyst who tracks the im­pact of in­for­ma­tion tech­nol­ogy on man­age­ment and mar­kets. He speaks fre­quently on cloud, data, and enterprise com­put­ing top­ics at in­dus­try events and on we­b­casts.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.