NTERNATIONALMONETARYFUNDArtificialIntelligence andWORKINGPAPERProductivity in EuropeFlorian Misch, Ben Park, Carlo Pizzinelli and Galen SheWP/25/67progressbytheauthor(s)and arepublished toelicit comments and to encourage debate.those ofthe author(s) and do not necessarilyrepresent the views of the IMF,its Executive Boardor IMF management2025APR
@2025IntemationalMonetary FundWP/25/67IMFWorkingPaperEuropeanDepartmentAxrificialIntelligence and Productivity in Europ,Prepared by Florian Misch', Ben Park, Carlo Pizinelli and GalenSher*tAuthorized for distribution by Stephan Danninger and Kristina KostialMonth2025IMF Working Papers describe research in progress by the author(s) and are published to elicitauthor(s) and do not necessarly represent the views of the IMF, its Executive Board, or IMF managemenhe discussion on AtifiallIntligence(A) oftencenters around its impact onproductfivity, butlence for Eurooe remains scarce.UsinatheAcemoalu(2024) approach we simulate theof Al adontion on total actor productivity for 31 European countries. We compile manyle in the near term. using reduced-formrelevant requlation that restricts Al usehe medium-termproductivity aainsfoent cumulatively overfivevears Whileha stmates sy cemogu 224)for theUS. Theys and are substantially larger in countries with higher incomesdataprivacycombinedcould reduce Europe'sproductiviygainsbyover30 percentifAlexposure were 50percent lower in tasks, occupations and sectors affected by requlationReCOMMENDEDCITATION:IMisch. F.. Park. B. Pizzinelli. C. and Sher. G. (2024). ArtificialIntelligence andProductity in Europe.IMFWorkingPaperWP/25/6;JEL Classification NumbersE24.J24.030.047Keywords:AtifialIntelicence:Producivity; Technology,RegulationFMisch@ifora:HPark2@imforg;CPizineli@imf orgAuthor'sE-Mail Address:GSher(@imf.org*Correspondingauthor. We thank Simon Burnel, EFna Dabla-Noris, Romain Duwal and David Kol frexcellet comments.Mahika Gandhiprovidedexcellent research assistance.OInternational Monetary Fund. Not for Redistribution,
WORKINGPAPERSArtificial Intelligence andProductivity in EuropePrepared by Florian Misch1, Ben Park, Carlo Pizzinelli and Galen SherCorresponding author.7MahikaGandhiprovidedexcellentresearchassistance OInternational Monetarv Fund Not for Redistribution,
IMFWORKINGPAPERSArtificialIntelligence and Productivity in EuropeContents1.Introduction......32.StylizedFacts63.Methodology......4.Results......84.1VaratoninMetium-enmeroneutit......4.2PreferredScena.i......124.3TheRoleofRequlation..155.ConclusionsandPolicyImpi*ti*.s...............Appendix1:SampleandCountry-SpecificDatal....................Appendix2:AlExposureofTasksandOccupationsAppendix3:AlAdoptionRate..........22Appendix4:LaborCostSavingsfromAl心Appendix5:National Occupation .Reulatio......29Appendix6:EUAlAct..30Appendix7:DataPrivacyLaws..31References....32INTERNATIONALMONETARYFUNDCInternationalMonetarvFund.NotforRedistribution
IMFWORKINGPAPERSArtificialIntelligence and Productivity in Europe1. IntroductionAtificial intlligence (Al) is often seen as a general-purpose tech nology that has the...
