62

TECHNOLOGYTRENDS

THATWILL—ANDWON’T—SHAPE2026

262TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

FROMOUR

CHIEFRESEARCHOFFICER

Thispastyearhasbeenshapedbyacomplexmixofmacrostabilizationandstructuralstrain.Inflationhascontinuedtoeaseinsomeregions,monetarypolicyisgraduallynormalizing,andsupplychainshavebecomemorepredictable.

However,geopoliticaltensionsstillweighheavilyonthemindsoftechnologydecisionmakers.Conflictzonesremainactiveandtariffpolicieshaveshakenupthestatusquo.Atthesametime,energyavailability,climatepressures,andskillsshortagescontinuetolimitglobaleconomicgrowth.Asaresult,thetechnologysectorhasshiftedfromcrisismodetobeingcautiouslyoptimistic.

Withinthisenvironment,2026willbedefinedbypragmatictransformation.AcrossthemarketsthatABIResearch

analyzes,aconsistentpatternisemerging:organizationswillinvestwheretechnologyreducesrisk,lowerscost,or

directlyboostsproductivity.Theappetiteforsweepingvisionarypromisesisfading.Enterpriseswantpracticalresults,quickwins,andsolutionsthataddresslong-standingoperationalchallenges.

ArtificialIntelligence(AI)willcontinuetobeakeytalkingpointintheboardroom.WhileGenerativeandAgentictoolswillcertainlyprogress,purchasingdecisionswillrevolvearoundtargetedworkflows,automationofroutinetasks,andstructuredorchestration.Newcloudmodelsarealsomaterializing,whichreflectenterprisedemandforsovereign,full-stacksolutions.Meanwhile,organizationsacrosslogistics,buildings,energy,andmanufacturingincreasinglyprioritizesolutionsthatenhancedecision-making,notmerelyvisibilityplatforms.

Takentogether,2026willbeayearofdisciplinedmodernization.Theorganizationsthatsucceedwillbethosethatreducecomplexity,delivermeasurablevalue,andaddressthemostimmediateoperationalchallenges.ABIResearchlooksforwardtohelpingclientsnavigatethistransformativeyear.

StuartCarlaw

RESEARCHSERVICEAREAS

5GDevices,Smartphones&Wearables 3

5G,6G&OpenRAN 4

AI&MachineLearning 5

Automotive 6

CitizenDigitalIdentity 7

Cloud 8

DigitalPaymentTechnologies 9

EnterpriseConnectivity 10

eSIM&SIMSolutions 11

FreightTransportation 12

Industrial&ManufacturingMarkets 13

Industrial&ManufacturingTechnologies 14

Industrial,Collaborative&

CommercialRobotics 15

IoTHardware 16

IoTMarkets 17

IoTNetworks&Platforms 18

LocationTechnologies 20

QuantumSafeTechnologies 21

RFID 22

SmartBuildings 23

SmartEnergy 24

SmartLiving&ConsumerTechnologies 25

SpaceTechnologies&Innovation 26

SupplyChainManagementSoftware 27

TelcoAI 28

TrustedDeviceSolutions 29

Warehousing&Fulfillment 30

Wi-Fi&WLANTechnologies&Market 31

Wi-Fi,Bluetooth&WirelessConnectivity 32

XRMarkets 33

XRTechnologies 34

362TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

5GDEVICES,SMARTPHONES&WEARABLES

SMARTWEARABLESALESWILLSTRENGTHEN.

1

WILL

HAPPEN

Thewearabledevicessectorrepresentsaconfluenceofminiaturizedtechnology,fashion,andpersonalaspirations/concerns.

Originally,thelineuponlyincludedwiredheadphonesandsmartphonescases,butsincethen,smartwatches,smartrings,smartclothing,andevenArtificialIntelligence(AI)-enabledglasseshavebeenaddedtothemix.

Oneofthemosticonicbastionsofthewearablesmarketisthesmartwatch.Seikomayhaveintroducedoneofthefirstsmart-watches,theSeikoRuputerin1998,butitwasnotuntiltheadventoftheAppleWatchin2015thatthesmartwatchmarkettookoff.The

COVID-19pandemictriggeredanupswellofsalesinsmartwatchesasenduserssoughtoutfitness,health,andwell-beingmonitors.

Thisledtoanoverhangofsalesthatdampenedunitshipmentsfrom2023to2025.However,ABIResearchanticipatesanupswinginsales(+6.5%inYear-over-Year(YoY)wearabledevicerevenue)thathasbeenledbythesmartwatchmarket.

Itisnotjuststepcountingandcaloriesburnedthataredrivingwearablesales,itisalsoenhancedhealthmonitoringfeatures.Forexample,theAppleSeries11Watchcannowwatchoutforhypertension,whiletheSamsungWatch8canmonitorantioxidantsinyourblood(akacarotenoidlevels).Severalsmartwatchesnowsupportsatelliteemergencymessaging.GooglerecentlyaddedthefeaturetoitsPixelline.

Smartringsarealsomakingtheirpresencefeltinthemarket.ABIResearchanticipatessmartringsalestogrow9.7%in2026

toUS$3.6billion.OuraandSamsunghavebeenprominentexponentsofthesmartringphenomenon,althoughRingConn,andUltrahumanarealsoclosingthegap.Improvedbatterylifeandsensorshavemadesmartringsmorediscreteandabletotakeonmoreaestheticfeatures.

Smartfeaturesandfunctionalitiesarealsomakingtheirwayintoclothing.Gesturerecognition,thermo-regulation,andinobtrusivemonitoringofhealthconditioningarejustsomeofthefunctionalitiesbeingaddedtosmartclothing.Smartclothinggenerated26.6millionitemsinsalesin2025andisexpectedtogrow15%in2026.

ON-DEVICEAIINSMARTPHONESWILLNOTBETHEDRIVEROFPREMIUMSMARTPHONEADOPTIONIN2026.

WON’THAPPEN

2

AIon-devicewasoneofthedefiningtopicsinsmartphonesin2025,withmobilevendorsaggressivelyrampinguptheirlaunchesforAI-enabledsmartphonesliketheGooglePixel10,theSamsungGalaxyAI-compatibledevices,AppleIntelligencecompatible

devices,andHuawei’snewestMate70lineofAIphones.However,whileAI-capablehardwareisalreadypresentinpremium

smartphonesshippingin2025,AIsoftwareinsmartphonesisyettocatchuptostandardsthatconsumersexpectorfindusefulforintegrationandimplementationintotheirlives.

MostsmartphonesshippedwithAIfeaturesnowcomewithsomeusecases,albeitalmostidenticaltoeachother.VendorscannotrelysolelyonfeatureslikeAI-enhancedphotography,AIphotoediting,AItextsummarization,livetranslation,andAI-poweredvoiceassistantstomarkettherevolutionarynatureofAIenablementinsmartphones.SustaininganddrivingdemandinnewAI-enabledsmartphoneswillrelyprimarilyonestablishingusefulconsumerexperiencesthatcanimpactandsimplifytheirdailylivesand

routines,notpurelyonhypecyclesandmarketing.Vendorsmustacknowledgethatthedriversinpremiumsmartphoneadoptionandshipmentsin2026willnotbeprimarilyreliantonthemarketingof“AI”asgamechangingifitsusecasesdonotevolveinto

moreconcretesoftwarethatchangesconsumers,dailyinteractionsandexperienceswiththeirsmartphones.

Currently,demandforAIonsmartphoneshasbeenlackluster,assmartphonevendorshavedonelittletoconvinceconsumersofthebenefitsofon-deviceAI,howprivacyisprioritized,andhowtointegrateAIeffectively,ratherthantreatingitasjusta“nicetohave”feature.Ifthisissueremainsunresolvedin2026,vendorsshouldnotexpectthaton-devicefeaturesinthenewestpremiumsmartphoneswillbetheprimarydriverofadoption,butratheraforgettableadd-onthatcomeswiththeincreasingpricetagof

premiumsmartphones.

462TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

5G,6G&OPENRAN

6GCAUTIOUSLYPROPOSESSERVICEINNOVATION,BUTTHEINDUSTRYISSTILLINSHOCKFROMTHE5GENTERPRISELETDOWN.

3

WILL

HAPPEN

In2026,discussionsabout6Gnetworkswillbecomeincreasinglycommonacrossthetelecomssector.Unlikethehardselling

thatoccurredfor5Gnetworks,themessagingaround6Gwillbemuchmoremeasuredandrealisticintone.Vendorsandserviceprovidersareawareofthecredibilitygapcreatedwhen5GEnterprisesolutionsdidnotproducetheexpectedtransformative

ReturnonInvestment(ROI)thatmanybusinessesexpected.Asaresult,6Gwillbemarketedasanevolutiontowardenhancedconnectivityinsteadofbeingbrandedasaparadigm-shiftingtechnology.

Initial6Gofferingswillfocusmoreoncreatingsmall-scaleserviceinnovationversuslarge-scaledisruption.Thefirstexperimentswillbemoreaboutusinghybridmodelsthatcombine5G-Advancedwithselectelementsof6Gtechnology,suchasintegratedsensingandArtificialIntelligence(AI)-drivenorchestration,andtargetingaspecificnichemarketofindustrialautomationandimmersive

ExtendedReality(XR)applications.Thesefirststepsfor6Gwillbepositionedasfuture-readysolutions,insteadofbeingthebasisforimmediatemonetizationopportunities.

Enterpriseswillstillbehesitanttoimplement6GpilotprogramsuntiltransparentbusinessmodelsareavailablethatshowthemhowtheywillachieveROI.Foroperators,software-definedupgrades,increasedspectrumefficiency,andcollaborativepartnershipsthatminimizeupfrontcapitalinvestmentswillbetoppriorities.Thethemewillcontinuetofocusonresponsibleinnovation,whilestandardsorganizationswillbeestablishingtechnicalspecificationsforvendorstodevelopproof-of-conceptdemonstrations

insteadofofferingbroad-basedtransformationpromisesassociatedwiththe5Gera.

AI-RANWILLNOTTAKEOFFIN2026ASOPERATORSDEMANDPROOFOVERPROMISES.

WON’THAPPEN

4

DiscussionssurroundingArtificialIntelligence-RadioAccessNetwork(AI-RAN)arebecomingincreasinglypopularinthe

telecommunicationssector,withithavingthepotentialtofacilitateautomation,improvenetworkperformance,andreduce

operatingcosts.Despitevendorinterestgainingmomentum,asevidencedbyamembershipgrowthrateoftheAI-RANAlliancefrom11initialmembersayearagotoover80today,ABIResearchdoesnotexpectAI-RANtoseeanymeaningfulcommercial

deploymentsin2026.Withoperatormembershipstallingat8members,thereisaclearoperatorreluctancetocommitfinanciallytoAI-RANbeforetheyseeclearevidenceofbothtechnologyperformanceandfinancialgain,especiallybecause5Gmonetizationremainsasignificanthurdleformanyofthem.

Additionally,thedisparitybetweentheperceivedadvantagesassociatedwithAI-RANandwhathasactuallybeendocumented

throughpracticalusecasesissimplytoogreatforserviceproviderstojustifycommittingtheircapitaltowardAI-RANdeploymentsatthistime.Currently,thereisnoindependentandvalidatedbenchmarkingprocess.Althoughtherehavebeensomepilot

programsrunbySoftBank,NVIDIA,andothersthatprovideinsightfulindicationsofpotentialcostreductionsandbetterefficiencies,noneofthoseprogramshavebeenindependentlyevaluatedorvalidatedbyathirdparty.

Architecturaluncertaintyaddstothecomplexityofthissituation.WhileGPU-basedsolutionscanprovidehigh-performanceforadvancedusecases,suchasmassiveMultipleInput,MultipleOutput(mMIMO),thereisconcernaboutGraphicsProcessingUnit(GPU)-basedsolutionsregardingtheirhigh-energyconsumption,beingproprietary,andthepotentialforvendorlock-in.While

CentralProcessingUnit(CPU)-basedsolutionsandcustomsiliconsolutionsaremuchmoreefficientforArtificialIntelligence(AI)workloads,neithercanprovidethelevelofaccelerationrequiredforexecutingthemostcomplexalgorithmscurrentlyusedforAI.Thislackofconsensuswillstalldecision-makingthroughout2026.

In2026,ABIResearchanticipatesonlygradualadvancementstowardimplementingAItechnologiesinapplicationsthatarenot

timesensitive,suchastrafficpredictionandanomalydetection,asthesetechnologieswillprovidesignificantbenefits,while

eliminatingtherisksofdeployingAI-RANsolutionscompletely.InorderforAI-RANtobecomemorethanjustabuzzword,the

telecomsindustrywillneedtostandardizemetrics,validateeconomicmodels,anddemonstrateinteroperabilitybetweendifferentarchitectures.Atbest,2026willbeayearwhenAI-RANislikelytoexperienceacontinuationofconservativetesting,withonly

modestdeploymentsoccurring.

562TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

AI&MACHINELEARNING

OPENSTANDARDSINAIDATACENTERSWILLPROLIFERATEANDGROWININFLUENCE.

5

WILL

HAPPEN

OpenstandardsthatenableinteroperabilitybetweenservervendorsandincreasethemodularityofArtificialIntelligence(AI)

clusterswillcontinuetogainimportancein2026.SuchstandardsareimportantforbuildingthenextgenerationofAIdatacentersbecausetheydismantleproprietaryecosystemsandfosteramorecompetitiveenvironment.Atthehardwarelevel,standardssetbytheOpenComputeProject(OCP)andtheUltraAcceleratorLink(UALink)enableamodular,multi-vendorapproach,allowing

operatorstointegratebest-of-breedaccelerators,networking,andstoragewithoutbeinglockedintoasinglevendor’sdesigns.

TheremitoftheleadingOCPbodywillalsoexpand,asseenattheirannualGlobalSummitinNovember2025,withitssphereofinfluencehavinggrownbeyondserverstoracks,clusters,andnowthelayoutofAIdatacenters.

Therewillalsobesignificantactivityintheinterconnectspace,asseveralnewstandardshaveemergedover2025,alongsidethe

maturationofexistingchallengerstoNVIDIA’sproprietaryNVLink,suchasUALink.ThisincludesBroadcom’sScale-UpEthernet

(SUE),unveiledinconjunctionwiththeTomahawkUltraswitchApplication-SpecificIntegratedCircuit(ASIC),whichcompeteswithInfiniBandandNVLink.OCPalsoannouncedtheEthernetforScale-UpNetworking(ESUN)workstream,whichwilladvanceEthernetinthescale-updomainandworkwithotheropenstandardssuchasUALink(spearheadedbyAMD)andtheUltraEthernet

Consortium(UEC)forscale-outacrosscomputingnotes.

THEMINTINGOFNEWAI-FIRSTNEOCLOUDSINANDNORTHAMERICA.

EUROPE

WON’THAPPEN

6

WithrumorsaboutAIclusterssittingidleforlengthsoftimeandneoclouds’laserfocusonefficiencyandutilization,itishighlyunlikelythatwewillseetheincorporationofnewneocloudsinEuropeorNorthAmerica,wherethemarkethasmatured

significantlyoverthelast12months.Thiscomeswiththeriskofmarketexitsandconsolidation,although,todate,thishasbeenlessthananticipated(NorthernDataGroup’sTaigaCloudandrumorsabouttheliquidationofGenesisCloudnotwithstanding).

Competitionistightasitstands,whichisalsoreflectedinthefallingneocloudon-demandGPUhourprices,whicharedistinctfromhyperscalerplatforms.Thisdynamicisalsoreflectedbyneoclouds’investmentsintheircloudsoftwareplatforms(forinferenceandtrainingworkloads),whichaimtoretaincustomersbybuildingacommercialmoataroundtheiroffering.

662TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

AUTOMOTIVE

CAMERA-BASEDDRIVERMONITORINGSYSTEMSWILLGROWRAPIDLYIN2026,DRIVENBYREGULATORYREQUIREMENTSINEUROPE.

7

WILL

HAPPEN

Camera-baseddrivermonitoringsystemswillexperiencerapidgrowthin2026,drivenbyabindingregulatoryrequirementintheEuropeanmarket.Historically,anichesystemhighlycorrelatedwiththeadoptionofsemi-autonomousdrivingsystems,camera-basedDriverMonitoringSystems(DMSs)areideallypositionedasthetechnologytofulfilltheAdvancedDriverDistractionWarning(ADDW)componentoftheEuropeanUnion’s(EU)GeneralSafetyRegulation(GSR2),directlytrackingthedriver’seyestomeasuredirectionofgaze.Whilethisrequirementfirstcameintoeffectin2024,itonlyappliedtonewlyintroducedmodels,withJuly2026thekeydatefromwhichallnewlyshippedvehicles,includingthosebasedonlegacyplatforms,mustfeatureADDW-fulfilling

technology.Asaresult,farmorevehicleswillshipin2026featuringcamera-basedDMSs,withABIResearchexpecting7millionDMS-equippedmodelstoshipin2026,withfurthergrowthexpectedin2027asthetimewindowexpandstocovereffectivelyallvehiclesshippinginEuropethatyear.

ThisiswelcomenewstoDMSsuppliersthathavelong-standingrelationshipswithEuropeanOriginalEquipmentManufacturers(OEMs),butwhichhavehadlittlechoicebuttowaitforbindingregulationtodeliversignificantvolumes.Withthegrowthimpactofthisregulationexpectedtobeexhaustedbytheendof2027,DMSsuppliersmustlooktoothermarketstosustaingrowth,withChinathenextfrontier,albeitwithadifferentsetofrequirementsslantedmoretowardfeaturerichness,aswellasanincreasinglycrowdedsetofdomesticsuppliersreadytodeliver.

THENEWVEHICLESALESMARKETWILLNOTBOUNCEBACKIN2026,ASAUTOMAKERSCONTINUETONAVIGATESUPPLYCHAINCHALLENGES.

WON’THAPPEN

8

In2025,theautomotiveindustryfaceditslargestsupplychaindisruptionsincethe2021semiconductorshortage,asaslewoftitfortattariffsbroughtchaostoautomotivesupplychains.Evenasindividualautomakerssoughttore-optimizetheirmanufacturingstrategies,adjustingthemanufacturinglocationsforkeymodelstominimizethetariffimpact,themarketasawholeexperiencedayearoftwohalves.Withmanyofthetariffspre-announcedaheadoftime,consumersthatanticipatedaloomingpricehitrushedouttobuynewcarsfromexistinginventory,withtheeffectofdeliveringabetterthanexpectedfirsthalf,althoughfalteringgrowthinthesecondhalfsuggeststhattheimpactofgrowingtradefrictionisstartingtobite.

Goinginto2026,automakersshouldnotanticipatearapidreturntogrowth.Evenasslowermovingautomakersmakethe

finaladjustmentstotheirsupplychainstrategies,theNexperiacrisisdemonstrateshowexposedtheautomotivemarketisto

disruptionswithinsupplierswithcircularsupplychainscrossingincreasinglyabrasivegeopoliticaldividinglines.AssomeOEMs

onceagainfindthemselveshavingtoreducemanufacturingoutputduetoalackofessential,mature-nodecomponents,itisclearthattheindustryhasyettoproperlyabsorbandrespondtothelessonslearnedfromthe2021semiconductorshortagecrisis,andservesasanimportantreminderthattariffsarefarfromtheonlysupplychaindisruptorthatautomakersmustnavigate.Overall,ABIResearchexpectsmodestgrowthof1.6%forpassengervehiclesin2026.

762TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

CITIZENDIGITALIDENTITY

DESPITEACLEARTRENDTOWARDDIGITIZATION,

PHYSICALSECURITYWILLREMAINPROMINENTTHROUGHOUT2026.

9

WILL

HAPPEN

DespitethecleartrendtowarddigitizationwithinthegovernmentIDmarket,physicalcredentialsremainacornerstoneforsecuringbothidentitiesandtransactions.

Digital-firstapproachesthathavebeenimplementedandscaledremainfewandfarbetweenonagloballevel,andalthoughthemarketwilllikelycontinuethetransitiontodigital-firstapproaches,thegovernmentIDmarketwillremainprimarilyphysicalfirstinnature,withmobileactingasacompaniontothephysical.

Physicalcredentialsare,therefore,notgoingawayanytimesoon,andinamarketwithlongdocumentlifespansandincreasinglysophisticatedandwell-resourcedcounterfeiters,continuedevolutioninphysicalsecurityfeaturestohelptacklenovelsubversionmethodsremainsessential.

Whetheraddressingthemarketfromapersonalizationandprintingperspectiveorsupplyingconsumablessuchasmaterials,films,andinks,enhancingandimprovingphysicalsecurityfeatureswillremainafocalpointthroughout2026.

LAGGINGSTANDARDIZATIONEFFORTSANDPERSISTENTTECHNICALANDPRACTICALOBSTACLESDELAYQUANTUMREADINESSWITHIN

GOVERNMENT-ISSUEDANDNATIONALIDCARDSBEYOND2026.

10

WON’THAPPEN

Theyear2025representedabreakoutyearforPost-QuantumCryptography(PQC),whichisincreasinglypenetratingdigitaltrustmarketsacrossverticalsandregions.However,hold-upsarestillexpectedregardingPQCpenetrationintotheidentitysmartcardmarket.Therearethreeprimaryobstaclesresponsibleforthecurrentstandstill,

1.AlackofstandardsfromtheInternationalCivilAviationOrganization(ICAO)regardingthePQCalgorithmstobeintegratedintopassports.

2.GiventhelargerkeysizeandincreasedcomputationaldemandsofPQC,thesealgorithmsareinherentlyincompatiblewiththememory-constrainednatureofnationalIDcardsandpassports,complicatingintegrationinawaythatbalanceshighcryptographicsecuritywiththedemandsforefficiencyandspeedofverificationwithinthebroaderidentityspace.

3.Interoperabilityrequirementsmeanthatmigrationacrosstheidentityecosystemmusthappeninsync,posinghighpracticalobstaclestoachievingquantumreadinessingovernment-issuedandnationalIDcards.Whileclosed-loopenvironments,

particularlythosewithhighsecurityassurancerequirementssuchasphysicalaccesscardsforgovernmentdepartments,arefreefromthesameinteroperabilityandstandardsconstraintsthatlimitPQCpassports,thesametechnicalconstraintsapply.

WhiletheannouncementofNationalInstituteofStandardsandTechnology(NIST)-approvedquantumalgorithmshasgivenchipvendors,semiconductormanufacturers,digitaltrustandPublicKeyInfrastructure(PKI)providers,cardissuers,andmanufacturersofsupportinghardware(e.g.,documentscanners,eGates,etc.)thegreenlighttobeginintegratingthequantumalgorithmsintosmartcards,withoutvertical-specificmandatesdictatingwhichalgorithmstointegrate,quantumreadinessisnotexpecteduntilpost-2026,despiterecentdevelopmentspertainingtodedicatedhardwareacceleratorsandPQCintegration.TheICAOisintheworkinggroupstageofideatingPQCstandards,yetgiventhetypicallyslow-movingnatureofthisprocess,real-worlddeploymentsremainawayoff.Lastly,demandstoimplementcrypto-agilityplacefurtherpressureonsizeandcomputationconstraints,

especiallygiventhelimitedcapacityforOver-the-Air(OTA)updateswithinsmartcardarchitectures.

862TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

CLOUD

IN2026,CLOUDSOVEREIGNTYTAKESOVERASTHEMARKETPIVOTSTOHARDWARETRANSPARENCYANDFULL-STACKCONTROL.

11IEN

Givenallofitsgeopoliticaltensions,theyear2025underlinedtheimportanceofsovereigndigitalinfrastructures.Shiftingexportcontrols,retaliatorysanctions,andunpredictablenationalsecuritydirectivesmadeitclearthatglobalclouddependenciesarenowgeopoliticalliabilities.Thestartingpointformostenterprisesin2025wasanarrowfocusondataresidencyanddataaccessibilitycontrols.In2026,thissovereigntyshiftwillreshapethecompetitivelandscapemoreprofoundly,asenterpriseswilldemandafullysovereigncloudsoftwarestackandnotbeobsessedwithdataresidencyanddataaccessibilityalone.

Enterpriseswillincreasinglyevaluatecloudsontransparency,control,andsupplychainclarity.ThiswillfavorproviderslikeSTACKITorNextGenCloudthatdeliveropencompute,predictableeconomics,andcontractualguaranteesagainstunilateralplatform

changes.AlternativecloudserviceprovidersthatarespecializinginprovidingGraphicsProcessingUnits(GPUs)inacloud-likebusinessmodel(socalledneoclouds)willpositionthemselvesascrediblealternativestoU.S.-headquarteredhyperscalers.

ForthebigU.S.-headquarteredhyperscalers,thismeanstheymustrampuptheirsovereigncloudofferingstoremaincompetitiveinanincreasinglyfragmentedmarket.Asenterprisesdemandafullysovereigncloudsoftwarestack,hyperscalerswillhavetoworkwithregionalplayers(liketelcosystemintegrators,neoclouds,andothers)eitherthroughacquisitionsorpartnershipstobeabletodeliverthedegreeofsovereigntythatenterprisesarerequesting.

12,NNEOCLOUDSWILLNOTUNLOCKENTERPRISEAIUSECASESATSCALE.

WiththeintensifiedhypearoundArtificialIntelligence(AI),2025sawtheemergenceofanewgroupofserviceprovidersthat

focusonprovidingGPUsinacloud-likeas-a-Service(aaS)businessmodel.Whiletheirbusinessmodelsoundsappealingatfirst

(providingGPUcapabilitiestoenterprises,whilefreeingthemfromthemassivebarriertoentrycalledCapitalExpenditure(CAPEX)),theseserviceprovidersabsorballthebusinesseconomicrisksthatcomewithjoiningtheAIhypetrain.

Illustratingthisistheexorbitantlyhighcapitalintensityofleadingneocloudproviders(200%to300%in2025),reflectingthe

enormousupfrontinvestmentsrequiredforGPUs,datacenterbuild-outs,powerandcoolingupgrades,andspecializednetworkinginfrastructure.Thislevelofcapitalburdenisfarbeyondtypicalcloudeconomicsandhighlightshowaggressivelytheseproviders

mustspendsimplytoremaincompetitiveintheAIinfrastructurerace.

Withthesedynamicsplayingout,neocloudswillbeforcedintoarapidandunavoidablerepositioning.Tomonetizetheirheavy

investments,theymustmoveaggressivelyclosertoenterpriseverticalsandpositionthemselvesasembeddedpartnerswithin

manufacturingfloors,clinicalsystems,financialoperations,andenergygrids.Currently,~95%ofenterpriseAIapplicationsfailtoadvancepasttheProofofConcept(PoC)stage(the“PoCtrap”),highlightinghowdesperatelytheyrequirehelpandguidancewhenitcomestowhichenterpriseAIapplicationstodeploy.Successfullytargetingenterprises,however,requiresadensenetworkofpartners,aswellasaproventrackrecordandasetofblueprintAIusecasesthatenterpriseownerscantakeandadapttotheirownworkflowsandrequirements.Neocloudproviderswillhavetobuildthislibraryofusecases(includingReturnonInvestment(ROI)andCostofInaction(COI)discussions),andthenecessarytrustlevels.

Itwilltakeyearstobuildthisreputation,realistically,sothesecloudserviceproviderswillhavetorelyonventurecapitaland

contractswithtechnologycompaniesandLargeLanguageModel(LLM)developersbeforetheycanmonetizetheirplatformwithadiversified,healthystreamofenterpriserevenue.

962TECHTRENDSTHATWILL—ANDWON,T—SHAPE2026

DIGITALPAYMENTTECHNOLOGIES

THEYEAR2026WILLMARKTHEFINALHOORAYFORTHEBIOMETRICPAYMENTCARD.

13IEN

Thebuzzaroundbiometricpaymentcardscontinuestofade,despiteheavyinvestmentinthetechnology.Barriersthatinclude

priceandcomplexenrollmentmechanismshavehinderedadoptionwithinthepaymentcardmarket.Vendorshavesubsequentlybegunrepositioningtowardnewendmarkets,notablyforaccesscontrolandcrypto-coldwallets,butwithnewpartnershipsandecosystemstosetup,progresshasbeenslow.

Zwipebecamethefirstcorporatevictim,announcingitsinsolvencyinearly2025,demonstratingtheclearchallengesinacceleratingcommercialadoption.In2026,wewillcontinuetoseeahandfulofbio