KPMGHealthcaresector.The FutureofAl andtheWorkforceimpactor GenerativeAlon HealtncareJolbsReleasina timetocareMay2025
Healthcare sector.The Future ofAland theWorkforceForewordrecent months, the suden and very public,our work provides valuable insight into the(Al) haeLTHT might focusefit.Akeyexampleative tasks, such asand drafting, withGen Allinistrativember of countries vylng to be globaiders n this tanstormational new techndlogy.time on patient,Overthe next threeelopour understandingitsmight be the greatest, how its of Gen Al working with cliniciansnd how it can helpbecificuse cases in specificwill create new opportunitiesprofessionals will increasinglyinterpretingGenAGenAecurity. Rolesnd regulatory compliance;a panaceafor theow and in the futureey will continues,whilst alsorformedby thet(LTHT).AsfarMichaelAllenIt to individual NHS providers,Partnertof Health and Socialss the world which, to alargewe havestarted todevelopourJenny Lewisy and where there is an opportunityDirector ofHumanResources,orkforce by deploying Gen Al to helpKPMGD2025KFMGLR.aUKIm
Healthcare sector. The Future of Al and the WorkforceContentsSection1:Section2:ExecutiveSummaryIntroductions and keyfindingsSection3:Section4:ThepotentialbenefitsofGenAlforDetalled findings: Opportunites for Genthe workforceAl in the NHSSection5Section6:Whatemplbyersshould considertoAdionshealtcaeorganisations nedto take tomaximisethe benefis ofusingAlintheirensue complance wth relevant legal andorganisationslegislative requirements,KPMG
Healthcare sector.The Future ofAlandtheWorkforceGlossarvoftermsRoboticProcessAutomation(RPA);:DeepLearning:RobolicProes.AutomaionivobesulebasedAtype of machine leaing that uses antifical euralinetworkswithmuliplelayers to learn complex patternss opening an email, dala enty, or copying& pastingfrom data.Itis particularly effective for tasks such asda Aland RPAcan wokin anderm o udomaleimage recogion, atu alanguaepocessing,more sophisticated and complextasksand machine translation.ArtificialIntelligence(Al):GenerativeAl:AbroadfeldencompasingtedevelomentofAtype ofAlthat cancreate newcontent, such as texter systems that can perform taskstypicallymages, music, or code, based on existing data.ing humaninteligence, such as leaming,Examplesincludelarge language moels LMG)and large image models (LIMs),LargeLanguageModels(LLMs):A type of Al that allows computers to learn from dataA type of Al trained on massive amounts of text datawithout beigexiliaity programed. I invovesenabling t to generate human-cqualitytext, translatealgorithms that can identify patterms in data andmakees, interpret information, write different kindspredictions based on those patterns,of content. and answer questions. LLMs are createdsing deep learningNeuralNetworks:elmageModels(LIMs):Atypeofmachinelearningmodelinspired by thestructure and function of the human brain.They consisAtypeofAltrainedonmassive amounts ofimagedala,of interconnected nodes or"neurons"that processenabling it to generate realistic images, edit existinginformation andl...
