PitchBook-新興空間簡(jiǎn)報(bào):信任與安全技術(shù)(英)_第1頁
PitchBook-新興空間簡(jiǎn)報(bào):信任與安全技術(shù)(英)_第2頁
PitchBook-新興空間簡(jiǎn)報(bào):信任與安全技術(shù)(英)_第3頁
PitchBook-新興空間簡(jiǎn)報(bào):信任與安全技術(shù)(英)_第4頁
PitchBook-新興空間簡(jiǎn)報(bào):信任與安全技術(shù)(英)_第5頁
已閱讀5頁,還剩7頁未讀 繼續(xù)免費(fèi)閱讀

下載本文檔

版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)

文檔簡(jiǎn)介

$48.5

$66.2

$192.7

$312.0

$324.3

$26.5

Overview

suiteoftechnologiesandmethods

designed

referstothe

Trust&safetytech

toprotectusersondigitalplatformsfromvariousonlineharms,includingmisinformation,fraud,andexploitation.Respondingtothedynamicnatureofdigitalthreats,thisfieldinvolvesimplementingsystemsforcontent

moderation,ageverification,andregulatorycompliancetoenhanceonlineandtrustworthiness.

safety

Background

digitallandscapeisincreasinglychallengedbythespreadofmisinformationanddisinformation,whichwasrecentlyidentifiedbyaWorldEconomicForum

surveyasthetopglobalriskoverthenexttwoyears,potentiallyfuelingsocietalpolarizationandconflictrisks.1Misinformationisfalseorinaccurateinformationthatisspreadwithouttheintenttomislead,whereasdisinformationisfalse

informationthatisdeliberatelydisseminatedwiththeintentiontodeceiveor

mislead.TheriseofgenerativeAIcontributestothesechallengesbyenabling

therapiddisseminationof“fakenews”andAI-generatedcontentlikedeepfakes,2therebycomplicatingcontentmoderation.Amidtechfirmlayoffs,trust&safetytechstartupsareemergingtooffersolutionsforfirmsteamstotacklethesegrowingonlinesafetyconcerns.

The

lackingin-house

1:“TheGlobalRisksReport2024,19thEdition,InsightReport,”WorldEconomicForum,January2024.

2:AccordingtoTheAssociatedPress,a“deepfake”is“amanipulatedvideoorotherdigitalrepresentation(ML)techniquesthatyieldseeminglyrealistic,butfabricated,imagesandsounds.”

producedbysophisticated

learning

machine

(pitchBooK

AliJavaheri

Analyst,EmergingTechnology

ali.javaheri@

EMERGINGSPACEBRIEF

Trust&SafetyTech

OriginallypublishedMarch28,2024

pbinstitutionalresearch@

Trendingcompanies

NewsGuard

)RealtyDefender

Trust&safetytechVC

dealactivity

48

39

35

31

5

201920202021202220232024*

Dealvalue($M)Dealcount

Source:PitchBook?Geography:Global*AsofMarch25,2024

Foraccesstomoreofthisdataand

PitchBook’sEmergingSpacestool,

accessafreetriallink

here

.

33

1

(pitchBooK

EmergingSpaceBrief:Trust&SafetyTech

Technologiesandprocesses

Trust&safetytechleveragesamultifacetedapproachtoprotectusersandplatformsfromonlinethreatsbyemployingacombinationofadvancedtechnologiesand

processes,including:

AI&ML:Centraltocontentmoderation,thesetechnologiesautomatethedetectionandmitigationofharmfulcontent,suchashatespeech,misinformation,andabusivematerial.AIandMLarepivotalinautomatedcontentanalysis,astheyanalyzetext,images,andvideostoefficientlyidentifyandsortproblematiccontent.

Naturallanguageprocessing(NLP):NLPplaysakeyroleinparsingandanalyzingtexttounderstanditsmeaning,particularlyfordetectinghatespeechandextremistcontent.However,itseffectivenessislimitedbythespecificityofitsapplicationtoparticulardomainsofobjectionablecontentandtheinherentchallengesingraspingthenuancesofhumanlanguage.

Digitalhashtechnology:ExemplifiedbyPhotoDNAandContentID,digitalhash

technologyiscrucialforcombatingchildsexualabusematerialandcopyright

infringement.Bygeneratingdigitalsignaturesforimagesandvideos,itenablesthe

rapididentificationandremovalofillicitcontent.Despiteitseffectiveness,challengesariseinmoderatingextremistcontentduetosubjectivedefinitionsandthepotentialforoverbroadcontenttakedowns.3

Imagerecognition:Leveragingdigitalhashtechnologies,imagerecognitiontoolsenhancetheabilitytoidentifyspecificobjectswithinimages,therebyaiding

themoderationprocess.Thesetoolsareinstrumentalinfilteringcontentbut

oftenrequirehumanoversightfornuanceddecision-makingduetolimitationsincontextualunderstanding.

Metadatafiltering:Thistechnologyusesthedescriptiveinformationwithindigitalfilestoidentifycontentthatmatchesspecificcriteria.Whileeffectiveinspottingcopyright-infringingmaterial,itsreliabilityischallengedbytheeaseofmetadatamanipulation,thushighlightingtheadaptivenatureofonlinethreats.

Ageverificationsystemsanddigitalidentityverification:Toensurecompliancewithagerestrictionsandtoprotectminors,thesetechnologiesverifyuseridentities,therebypreventingimpersonationandensuringusersmeetplatformagerequirements.

Frauddetectionsystems:Utilizingpatternrecognitionandanomalydetection,

thesesystemsarecrucialforidentifyingandpreventingfraudulentactivities,thussafeguardinguserdataandplatformintegrity.

Regulatorycompliancesolutions:Thesetoolsassistplatformsinadheringtoacomplexlandscapeoflocalandglobalregulationsrelatedtoonlinesafety,dataprotection,and

copyrightlaws,thusensuringthatplatformsoperatewithinlegalframeworks.

3:“HowAutomatedToolsAreUsedintheContentModerationProcess,”NewAmerica,n.d.,accessedMarch20,2024.

2

(pitchBooK

EmergingSpaceBrief:Trust&SafetyTech

Userbehaviormonitoring:Thesetoolsanalyzeuserinteractionstoidentifysignsofabusiveorfraudulentbehavior,whichiscrucialformaintainingplatformintegrity.

Contentmoderation:Acombinationofautomatedtoolsandhumanoversightensuresthatcontentadherestocommunitystandardsandregulations,thusbalancingthe

needforrapidresponsewiththerequirementfornuancedjudgment.

Applications

Trust&safetytechnologiesplayavitalroleinvarioussectorsbyenhancinguser

safety,maintainingintegrity,andcomplyingwithregulations.Thesetechnologies

areappliedacrossdifferentplatforms:Insocialmedia,theyfilteroutinappropriate

contentandmisinformationbyusingcontentanalysis,imagerecognition,andNLP.

E-commerceplatformsandonlinemarketplacesutilizefrauddetection,digitalidentityverification,andcompliancesolutionstosecuretransactionsandprotectuser

data.Onlinegamingandesportsplatformsintegrateageverificationandbehaviormonitoringtocreatefairandsecureenvironments.Financialservicesandfintech

sectorsemployalgorithmsforfraudprevention,identityverification,andadherencetoregulatorystandards.Educationtechnologyplatformsimplementcontent

moderationandageverificationtoensureasafelearningspace,whiledatingapps

usesimilartechnologiesforuserprotectionagainstharassmentandfraud.Streaming

servicesrelyondigitalhashtechnologyandmetadatafilteringforcopyright

compliance,whereasonlineforumsandcollaborativeplatformsleverageAI-driven

toolsfordiscussionmoderationandcyberbullyingprevention.Governmentandpublicservicesalsoadoptthesetechnologiestosecuredigitalservicesandprotectsensitivedata.Asdigitalenvironmentsevolve,theclearimportanceoftrust&safetytechin

safeguardingusersandupholdingplatformintegrityacrossindustrieshighlightsanongoingneedforinnovativeandadaptivesolutions.

Limitations

Despitethecriticalroletrust&safetytechnologiesplayindigitalplatforms,they

faceseverallimitations.Aprimarychallengeisthedifficultyofaccuratelydetecting

andmoderatingcontentwithoutinfringingonfreedomofexpressionorcausing

unintendedcensorship,especiallygiventhesubtlenuancesoflanguageandculturalcontext.Automatedsystems,whileefficient,oftenlackthesophisticationtofully

understandthecomplexitiesofhumancommunication,thusleadingtoerrorsin

contentmoderation.Additionally,thesetechnologiescanstrugglewiththerapid

evolutionofonlinethreatsandtheingenuityofmaliciousactorswhocontinuously

devisenewwaystocircumventsafetymeasures.TherelianceonlargedatasetsforMLmodelsalsoraisesconcernsaboutprivacyandthepotentialforbias,whichcanskewtheeffectivenessofthesetechnologiesacrossdifferentdemographics.Furthermore,theimplementationoftrust&safetytechrequiressignificantinvestment,whichcanbeabarrierforsmallerplatforms.Lastly,theeffectivenessofthesetechnologiescanbelimitedbythelackoftransparencyandaccountabilityinhowtheyaredeployed,

makingitchallengingtoassesstheirtrueimpactondigitalsafetyandtrust.

3

(pitchBooK

EmergingSpaceBrief:Trust&SafetyTech

Recentdealactivityandmarketoutlook

Thetrust&safetytechsector,whilerelativelynascent,hasshownpromisingdeal

activity,indicatingrobustinvestorinterestandthepotentialforsignificantgrowth.Inrecentdevelopments,RealityDefender,whichspecializesinthedetectionofdeepfakesandAI-generatedcontent,secureda$19.4millionSeriesAinSeptember2023led

byDCVC.Thisinfusionofcapitalunderscoresthemarket’sdemandforadvanced

solutionsincombatingsophisticatedonlinethreats.Similarly,UK-basedLogically,

whichoffersaplatformforidentifyingandrespondingtomisinformation,raised$11.6millioninalate-stageroundinAugust2023,highlightingtheurgencytoeffectivelyaddressthespreadofmisinformation.

Intherealmofbiometricverification,Paravision,astartupfocusingonfacial

recognitiontechnologyforuserauthentication,raisedanotable$38.8millionSeriesAinJuly2023withbackingfromHICOInvestmentGroupandJ2Ventures.Moreover,Unitary,whichdevelopscomputervisionmodelstoflagharmfulcontentonline,

successfullyraised$24.6millioninaSeriesAinJuly2023ledbyCreandumandsupportedbyPluralPlatformandPaladinCapitalGroup.

Formoreonventurecapitalactivityin

digitalriskprotection,refertoour

Q42023

InformationSecurityReport

.

Thesefundingroundsnotonlydemonstratethesolidstreamofinvestorconfidencebutalsosignalthesector’sreadinesstoscaleandinnovate.Astrust&safetytech

continuestoevolve,itispositionedtoaddressamultibillion-dollarmarketinthe

comingyears,drivenbytheincreasingnecessityforplatformstoensurethedigitalsafetyandintegrityoftheircontentanduserinteractions.Thisburgeoningsector

standsatacriticaljunctionbetweentechnologyandsocietalneed,withthepotentialtodeliverimpactfulsolutionsacrossvariousdigitallandscapes.

4

EmergingSpaceBrief:Trust&SafetyTech

Quantitativeperspective*

198

426

companies

deals

46

deals(TTM)

-14.8%YoY

$4.2M

mediandealvalue

(TTM)

40.0%YoY

Toptrust&safetytechcompaniesbytotalraised*

716

investors

$45.6M

medianpost-money

valuation(TTM)

187.0%YoY

$7.6B

capitalinvested

$1.4B

capitalinvested

(TTM)

64.6%YoY

*AsofMarch25,2024

CompanyTotalraised($M)Lastfinancingsize($M)LastfinancingdateLastfinancingdealtypeHQlocationYearfounded

TaskUs$882.7$460.0September7,2022DebtrefinancingNewBraunfels,US2008

NordSecurity$200.0$100.0September28,2023Late-stageVCAmsterdam,Netherlands2012

Yoti$122.3$9.5December22,2023Late-stageVCLondon,UK2014

ActiveFence$98.4N/AMarch1,2023Late-stageVCTelAviv,Israel2018

Paravision$90.1$38.8July28,2023Late-stageVCSanFrancisco,US2013

VideoVerse$83.9$45.0December18,2023PEgrowth/expansionLosAngeles,US2016

Logically$50.3$11.6August11,2023Late-stageVCBrighouse,UK2017

SpectrumLabs$45.0N/ASeptember6,2023M&AMiami,US2016

Truepic$36.4$26.6August6,2021Late-stageVCSanDiego,US2016

Modulate$36.0N/ANovember1,2023Late-stageVCSomerville,US2017

Source:PitchBook?Geography:Global?*AsofMarch25,2024

Toptrust&safetytechcompaniesbyExitPredictorOpportunityScore*

CompanyOpportunityScoreSuccessprobabilityM&AprobabilityIPOprobabilityTotalraised($M)HQlocationYearfounded

NordSecurity9998%19%79%$200.0Amsterdam,Netherlands2012

TrustLab9793%91%2%$22.9PaloAlto,US2019

Logically9586%80%6%$50.3Brighouse,UK2017

Pasabi9386%85%1%$3.1Edinburgh,UK2015

RealityDefender9285%84%1%$22.4NewYork,US2018

Modulate8779%78%1%$36.0Somerville,US2017

VineSight8678%77%1%$4.1TelAviv,Israel2016

Unitary8376%75%1%$26.6London,UK2019

Nisos8192%91%1%$33.7Arlington,US2015

VdoCipher8073%72%1%$0.2Gurgaon,India2015

Source:PitchBook?Geography:Global?*AsofMarch25,2024

Note:Probabilitydataisbasedon

PitchBookVCExitPredictormethodology

.

5

EmergingSpaceBrief:Trust&SafetyTech

Toptrust&safetytechcompaniesbyactivepatents*

Company

ActivepatentdocumentsTotalraised($M)HQlocationYearfounded

Yoti

84

$122.3

London,UK

2014

Truepic

37

$36.4

SanDiego,US

2016

ImageAnalyzer

20

N/A

Quedgeley,UK

2005

Zoomph

13

$6.1

Reston,US

N/A

Modulate

12

$36.0

Somerville,US

2017

ZapFraud

12

N/A

PortolaValley,US

2014

TwoHat

9

$9.6

Kelowna,Canada

2012

Attestiv

8

$4.6

Lehi,US

2018

SafeCreative

7

N/A

Zaragoza,Spain

2007

Source:PitchBook?

Geography:Global?*AsofMarch25,2024

Toptrust&safety

techinvestors*

Investor

Investments

Primaryinvestortype

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。

最新文檔

評(píng)論

0/150

提交評(píng)論