SubmissiontoUNSpecialRapporteurforFreedomofExpressionandOpinionDavidKaye:ContentRegulationintheDigitalAgeRankingDigitalRightsDecember15,2017byIlanaUllman,PolicyandCommunicationsAnalyst
LauraReed,SeniorResearchAnalystandCoordinator
RebeccaMacKinnon,Director
TableofContents
ExecutiveSummary 2
AboutRankingDigitalRights 3
Introduction 4
1.Transparency 6
2.Impactassessment 12
3.GrievanceandRemedy 15
Conclusion 18
Recommendations 19
Appendix:Fulltextofrelevant2017CorporateAccountabilityIndexindicators 21
2
ExecutiveSummary
Internetplatformsaroundtheworldareunderincreasingpressurefromgovernmentsandmanyotherstakeholderstoremoveobjectionablecontent.Frommisinformationtohatespeechtoextremistcontenttoharassmentandabuse,theseplatformshavebecomegatekeepersofwhatinformationthepubliccanaccessandshare.Yet,despitetheimportantroletheyplayinmediatingpublicdiscourse,anddespiteprogressbysomecompaniesinrecentyearsindisclosingpoliciesandactionsrelatedtogovernmentrequests,theprocessofpolicingcontentoninternetplatformsremainsunacceptablyopaque.Asaresult,usersofinternetplatformscannotadequatelyunderstandhowtheironlineinformationenvironmentisbeinggovernedandshaped,bywhom,underwhatauthorities,forwhatreason.Whentransparencyaroundthepolicingofonlinespeechisinadequate,peopledonotknowwhotoholdaccountablewheninfringementsoftheirexpressionrightsoccur.Thissituationisexacerbatedbythefactthatsomeoftheworld’smostpowerfulinternetplatformsdonotconductsystematicimpactassessmentsofhowtheirtermsofservicepoliciesandenforcementmechanismsaffectusers’rights.Furthermore,grievanceandremedymechanismsforuserstoreportandobtainredresswhentheirexpressionrightsareinfringedarewoefullyinadequate.InthissubmissiontotheUNSpecialRapporteuronFreedomonExpressionandOpinionDavidKayeforhisupcomingreportoncontentregulationinthedigitalage,RankingDigitalRightsproposesthefollowingrecommendationsforcompaniesandgovernments:
1. Increasetransparencyofhowlawsgoverningonlinecontentareenforcedviainternetintermediariesandhowdecisionstorestrictcontentarebeingmadeandcarriedout.Companiesshoulddisclosepoliciesfordecisionmakingregardingcontentrestrictions,whetherattherequestofgovernments,privateactors,orcarriedoutatthecompany’sowninitiativetoenforceitstermsofservice.Theyshouldalsodisclosedataonthevolumeandnatureofcontentbeingrestrictedorremovedforthefullrangeofreasonsthatresultinrestriction.Governmentsmustencourageifnotrequiresuchtransparencyandmatchitwithtransparencyoftheirownregardingdemands–directaswellasindirect–thattheyplaceuponcompaniestorestrictcontent.
2. Broadenimpactassessmentandhumanrightsduediligenceinrelationtothe
regulationandprivatepolicingofcontent.Companiesmustconducthumanrightsimpactassessmentsthatexaminepoliciesandmechanismsforidentifyingandrestrictingcontent,includingtermsofserviceenforcementandprivateflaggingmechanisms.Theymustdisclosehowsuchassessmentsareusedidentifyandmitigateanynegativeimpactonfreedomofexpressionthatmaybecausedbythesepoliciesandmechanisms.Governmentsshouldalsoassessexistingandproposedlawsregulating
3
contentoninternetplatformstoensurethattheydonotresultinincreasedinfringementofusers’freedomofexpressionrights.
3. Establishandsupporteffectivegrievanceandremedymechanismstoaddress
infringementsofinternetusers’freedomofexpressionrights.Whencontentiserroneouslyremovedoralaworpolicyismisinterpretedinamannerthatresultsinthecensorshipofspeechthatshouldbeprotectedunderinternationalhumanrightslaw,effectivegrievanceandremedymechanismsareessentialtomitigatingharm.Adequatemechanismsarepresentlylackingontheworld’slargestandmostpowerfulinternetplatforms.Governmentsseekingincreasedpolicingofextremistandviolentcontentbyplatformsshouldnotonlysupportbutparticipateinthedevelopmentofeffectivegrievanceandremedymechanisms.
AboutRankingDigitalRights
RankingDigitalRights(RDR)isanon-profitresearchinitiativehousedatNewAmerica’sOpenTechnologyInstitute,workingwithaninternationalnetworkofpartnerstosetglobalstandardsforhowcompaniesintheinformationandcommunicationstechnology(ICT)sectorshouldrespectfreedomofexpressionandprivacy.Formoreabouttheprojectsee:https://rankingdigitalrights.orgIn2015,RDRlauncheditsinauguralCorporateAccountabilityIndexwhichevaluated16companiesbasedon31indicatorsfocusedoncorporatedisclosureofpoliciesandpracticesthataffectusers’freedomofexpressionandprivacy.InMarch2017,RankingDigitalRightsreleasedthesecondeditionofitsCorporateAccountabilityIndex,whichranked22companiesaccordingtoanexpandedlistof35indicators.ThefullIndexresults,includingthereportandrawdataforresearcherstodownloadanduse,canbefoundat:https://rankingdigitalrights.org/index2017/The2017CorporateAccountabilityIndexevaluated22internet,mobile,andtelecommunicationscompanieson35indicatorsassessingcompanies’publicdisclosuresandcommitmentsinthreecategories:governance,freedomofexpression,andprivacy.Ofthe22companiesevaluated,tenprovidesearchand/orsocialplatforms(allservicesforwhichthecompanieswereevaluated,includingnon-platformservices,arelistedbelow):
● Baidu(China)—BaiduSearch,BaiduCloud,BaiduPostBar ● Facebook(US)—Facebook,Instagram,WhatsApp,Messenger ● Mail.Ru(Russia)—VKontakte,Mail.Ruemail,Mail.RuAgent
4
● Microsoft(US)—Bing,Outlook.com,Skype ● Kakao(SouthKorea)—DaumSearch,DaumMail,KakaoTalk ● Google(US)—Search,Gmail,YouTube,Androidmobileecosystem ● Tencent(China)—QZone,QQ,WeChat ● Twitter(US)—Twitter,Periscope,Vine ● Yahoo(US)—YahooMail,Flickr,Tumblr ● Yandex(Russia)—YandexMail,YandexSearch,YandexDisk(cloudstorage)
Thefullmethodology,alongwithresearchguidance,canbefoundhere:https://rankingdigitalrights.org/2017-indicators/Thissubmissioncoversresultsfromfiveindicators(F3-F7)thatevaluatescorporatetransparencyaboutarangeofcompanyactionsthataffectusers’freedomofexpression,plustwoindicatorsthatevaluatedifferentaspectsofgovernance:G4evaluatesdisclosureaboutimpactassessments,andG6evaluatesdisclosureofgrievanceandremedymechanisms.ThethirdCorporateAccountabilityIndex,whichcoversthesame22companieswiththesamemethodology,willbereleasedinApril2018.Whileresearchforthe2018Indexisalreadyunderway,itisfarfromfinalasoftheDecember2017deadlineforthissubmission.Theinformationinthissubmissionisthereforebasedonthe2017CorporateAccountabilityIndex,publishedinMarch2017.
Introduction
Companiesaresubjecttogovernmentregulationandalsoself-regulate.Self-regulationissometimescarriedoutunilaterallybyacompany,andissometimescarriedoutinconjunctionwithotherstakeholderswithgovernmentsupportasanalternativetodirectregulation(sometimesreferredtoas“co-regulation”).1Ofthe35indicatorsusedtoevaluatecompaniesinthe2017RankingDigitalRightsCorporateAccountabilityIndex,tenaredirectlyrelevanttothequestionofhowcontentregulationaffectsfreedomofexpressiononinternetplatforms.Theresultsofthe2017Indexshowthatcompaniesdisclosethegreatestamountofinformationabouttheircommitments,policies,andprocessesinresponsetodirectgovernmentdemandstoremoveorrestrictcontentordeactivateuseraccounts.Theydisclosetheleastamountofinformationabouthowprivaterulesandmechanismsforself-andco-regulationareformulatedandcarriedout.TheresultsofRDR’s2017IndexalsoshowthatwhilemanycompaniesintheIndexconductassessmentsofhowgovernmentdemandsaffecttheirusers’rights,fewcompaniesappeartoconductimpactassessmentstoidentifyhowtheirownprivateenforcementpoliciesandpracticesaffectthefreedomofexpressionofusersaroundtheworld. 1Seep.54,http://unesdoc.unesco.org/images/0023/002311/231162e.pdf
5
RDR’sevaluationalsorevealsthatgrievanceandremedymechanismsofferedbycompaniestousersarefarfromadequate.ConcernsthatglobalinternetplatformsarecensoringspeechthatshouldbeprotectedunderinternationalhumanrightslawintheireffortstocomplywithgrowinggovernmentregulatorypressuresarereflectedinrecentdebatesandcritiquesofGermany’s“NetzDG”law,whichcameintoforceinOctober2017.2Thelawimposessignificantfinesonsocialmediaplatformsthatdonotremovehatespeechcontent,whichisillegalunderGermanlaw,within24hours.Companiesthatrepeatedlyfailtodeletethisillegalcontentinatimelymannercouldfacefinesofupto€50million.3Thefullimpactofthelawremainstobeseen,butisevenmoreimportanttounderstandinthecontextofothergovernments,suchasFranceandtheUKconsideringnewfinesforsocialmediasitesthatdonotremoveextremistcontent,andincreasingpublicpressureintheU.S.toclampdownonwhitesupremacistcontent.4Hastyandbluntapplicationofenforcementmechanismsbeingusedinresponsetosuchlawshasresultedinrestrictionofprotectedspeech.Onewell-knownexampleistherecentdeletionofSyrianoppositionvideosonYoutube.5Suchcaseshighlightwhygreatertransparency,impactassessmentandremedyarevitalasgovernmentsandcompaniesstrugglewiththequestionofhowtoprotectthepublicandusersfromhatespeechandextremistattackswithoutviolatinginternetusers’expressionrights.Ascompaniesstrengthenpoliciesandbeefuptechnicalmechanismstopolicesuchcontent,lackoftransparencyaboutthesemechanisms,oraboutthevolumeandnatureofcontentbeingremovedbythem,makesitimpossibleforstakeholderstoknowwhetherthesepoliciesandmechanismsareachievingtheirintendedpurposes.Lackofimpactassessmentmeansthatcompaniesmaynotthemselveshaveaclearunderstandingaboutthepotentialcollateraldamagetheirpoliciesandmechanismsmayinflictandhowtomitigatenegativeimpactsonusers’freedomofexpression.Lackofadequategrievanceandremedymechanismsmeansthatwhenactivistsorjournalistsorothersexercisingtheirfreeexpressionrightsaresilencedbyaplatformusingmechanismsintendedtosilencehatespeechandextremism,theydonothavereliablerecoursetoaprocessforhavingtheircasereconsideredandcontentreinstated.6
2https://edri.org/eu-action-needed-german-netzdg-draft-threatens-freedomofexpression/3 https://qz.com/1090825/germanys-new-social-media-law-analysis-facebook-twitter-youtube-to-remove-hate-speech-in-24-hours-or-face-fines/4 https://www.theverge.com/2017/6/13/15790034/france-uk-social-media-fine-terrorism-may-macronandhttp://thehill.com/policy/technology/347173-tech-companies-crack-down-on-hate-speech-after-charlottesville5 https://www.nytimes.com/2017/08/22/world/middleeast/syria-youtube-videos-isis.html6https://theintercept.com/2017/11/02/war-crimes-youtube-facebook-syria-rohingya/
6
1.Transparency
FindingsfromRDR’s2017Indexhighlightspecificareasofweaknessaswellassomeemergingpractices.EightindicatorsintheIndexevaluatecompanydisclosuresabouttheirpoliciesandmechanismsrelatedtogovernmentrequestsandlegalcomplianceaswellastheenforcementofprivaterules,setbythecompany,aboutwhattypesofspeechandactivityarepermissible.7The2017Indexresultsspotlightkeyareasinwhichinternetplatformscanimprovetransparencyabouttheircontentmoderationpoliciesandpractices.Wefoundthatcompaniesoveralllacktransparencyabouttheirpoliciesaffectingusers’freedomofexpression—andinparticularaboutwhattypesofcontentareprohibitedandwhattheirprocessisforenforcingtheserules.Companiesalsotendedtodisclosemoreinformationaboutrequeststheyreceivefromgovernmentsandprivatepartiestorestrictordeletecontentordeactivateaccountsthanaboutactionscompaniesthemselvestooktoenforcetermsofservice.Companiesalsolackeddisclosureofwhethertheynotifyuserswhentheyrestrictcontentoraccounts.1.1Transparencyaboutcompanyactionsthataffectfreedomofexpressionlagsbehindtransparencyaboutactionsthataffectprivacy.Transparencyreportsareonewayforcompaniestoregularlypublishdataaboutthirdpartyrequeststheyreceiveandcomplywith,forbothcontentremovalsanduserinformation,andareincreasinglyacommonpractice.8Althoughmorecompanieshavebegunissuingtransparencyreports,companiestendtoreportmoreinformationaboutrequeststheyreceivetoshareuserdata(affectingprivacy)thantheydoforactionsthataffectfreedomofexpression,suchascontentrestrictionandremovaloraccountdeactivation.Figure1onthenextpagecomparescompanyscoresontransparencyreportingaboutrequeststhattheyreceivetoshareuserinformation(darkblue)9versustransparencyreportingaboutrequeststheyreceivetorestrictorremovecontent,ordeactivateaccounts(lightblue)10.
7SeeF1-F8athttps://rankingdigitalrights.org/2017-indicators/#F8 https://www.newamerica.org/in-depth/getting-internet-companies-do-right-thing/case-study-3-transparency-reporting/9 Seehttps://rankingdigitalrights.org/index2017/indicators/#P1110Seehttps://rankingdigitalrights.org/index2017/indicators/#F6andhttps://rankingdigitalrights.org/index2017/indicators/#F7
7
1.2Companiesvarywidelyintheirdisclosuresaboutgovernmentrequestsversusdisclosuresaboutrequestsfromprivateparties,includingrequestsrelatedtotermsofserviceviolations.Allplatformcompanieshaveroomtoimprovetheirtransparencyaboutrequestsaffectingfreedomofexpression,andshouldbejustastransparentaboutgovernmentrequeststoremovecontentastheyarewithprivateones(andviceversa).Currently,usersareleftwithanincompletepictureofthescopeandpotentialimpactthatgovernmentandprivaterequestsmayhaveontheirspeech.NotethatRDRdefines“governmentrequests”forcontentremovalsasrequestsoriginatingfromgovernmentministriesoragencies,lawenforcement,orcourtordersincriminalandcivilcases.RDRdefines“privaterequests”asthosemadebyanypersonorentitynotactingunderdirectgovernmentalorcourtauthority.Examplesofprivaterequestsincluderequestsfromaself-regulatorybodysuchastheUK’sInternetWatchFoundation,oranotice-and-takedownsystem,suchastheU.S.DigitalMillenniumCopyrightAct.
8
Figure2belowcomparescompanyscoresondatadisclosedaboutgovernmentrequests(F6)11withtheirscoresondatadisclosedaboutprivaterequests(F7).12
Disclosureaboutprivaterequestsforcontentrestrictionisalsoimportantformonitoringthefullimpactofgovernmentrequestsforcontentrestriction,giventhatgovernmentsoften
11Seehttps://rankingdigitalrights.org/index2017/indicators/#F612Seehttps://rankingdigitalrights.org/index2017/indicators/#F7
9
delegatetake-downrequestsandthereportingofterms-of-serviceviolationstoprivateparties.Forexample,therehavebeendocumentedcasesofcopyrightenforcementmechanismsbeingabusedbygovernments,suchasEcuadorPresidentRafaelCorrea,whousedmillionsofdollarsofpublicfundstohireaforeigncompanytohelpdeleteinformationcriticalofhimfromsitesincludingYouTube,Facebook,Vimeo,andDailymotion.13OfparticularnoteisFacebook’sminimaldisclosureofdatarelatedtoanytypeofprivaterequestsforcontentrestriction(F7).Itsscoreonthisindicator,6%outofapossible100%,reflectedtheleastamountofdisclosureincomparisonwithitspeers,exceptforthreecompaniesheadquarteredinmuchmorerepressivespeechenvironments(ChinaandRussia).Figure3belowcomparesFacebook’srelativelystrongperformanceindisclosingdataaboutgovernmentrequestsforuserdata,incontrasttoitspoorperformanceindisclosingcompanyactionsthataffectusers’freedomofexpression.
Allplatformcompanieshaveroomtoimprovetheirtransparencyaboutrequeststheyreceiveandhowthoserequestsarehandled.Theyshouldbejustastransparentaboutgovernmentrequeststoremovecontentastheyarewithprivateones(andviceversa).Currently,usersareleftwithanincompletepictureofthescopeandpotentialimpactthatgovernmentandprivate
13 https://www.buzzfeed.com/jamesball/ecuadors-president-used-millions-of-dollars-of-public-funds
10
requeststoremovecontentmayhaveontheirspeech,andmorebroadlyontheglobalflowofinformationonline.1.3Companiesdisclosemoreinformationaboutprocessesthandataaboutthevolumeandnatureofrestrictionsandremovals.Therecontinuestobeagapinwhatcompaniesdiscloseabouttheirroleincensoringonlinecontent,withcompaniesdisclosingmoreabouttheirprocessesforrespondingtogovernmentandprivaterequeststorestrictcontentandaccountsthantheydoaboutthenumberofrequeststheyreceiveandwithwhichtheycomply.Thismakesitunclearhowtheseprocessesareappliedinpractice,anddifficulttodeterminethescopeandpotentialimpactofthesecontentrestrictionsonfreedomofexpression.Inlightoftherecentincreaseingovernmentpressuretoremoveextremist,hatespeech,andotherobjectionablecategoriesofcontent,itisparticularlyimportantthatcompaniespublishdataonthecontentrestrictionrequeststheyreceiveandcomplywith,sothatadvocatesandthepubliccandetermineifcontentremovalrequestsareontherise,andwhetherornotthecompanyispushingbackagainstthem.Inaddition,whengovernmentsboastofincreasedcooperationwithplatformcompanies,thisdatamayalsooffersomeinsighttohelpdeterminewhethertheseclaimshavemerit,oraremerelymorerhetoric.Forexample,inApril2017,theVietnamesegovernmentreportedithadreachedanagreementwithFacebookforittocensorcontentthatviolateslocallawsorpostswith“fakecontent”aboutgovernmentofficials.14VietnameseofficialsclaimedthatFacebookhadagreedtosetupadirectchannelofcommunicationwiththegovernmenttofacilitatetheserequests,althoughFacebookstatedthatitsprocessforreceivingandrespondingtogovernmentrequestsisconsistentacrossjurisdictions.15Notably,initsGovernmentRequestsReport,Facebookdisclosesonlythenumberofpiecesofcontentitrestrictedduetogovernmentrequests,percountry.16Itdoesnotdisclosethenumberofrequestsitreceived,makingitimpossibletotelltowhatdegreethecompanyiscomplyingwith,orperhapspushingbackagainst,governmentrequestsitreceivestocensorcontentonitsplatform.Thisgapbetweendisclosureofcontentpoliciesandhowtheyareenforcedinpracticeisparticularlynoticeableinthecaseoftermsofserviceenforcement,forwhichthedifferenceincompanyperformanceamongthetwoindicatorsisparticularlypronounced. 14 https://www.reuters.com/article/us-facebook-vietnam/vietnam-says-facebook-commits-to-preventing-offensive-content-idUSKBN17T0A015 https://www.reuters.com/article/us-facebook-vietnam/vietnam-says-facebook-commits-to-preventing-offensive-content-idUSKBN17T0A016 https://govtrequests.facebook.com/
11
AsFigure4belowillustrates,whileallcompaniesdiscloseinformationaboutprocesses,onlythreecompanieswerefoundtohavedisclosedanydataaboutthevolumeandnatureofcontenttheyremovedattheirowninitiativewhenenforcingtheirtermsofservice:
Ofthethreecompanies—Google,Microsoft,andTwitter—thatdisclosedanydataaboutcontenttheyrestrictedasaresultofenforcingtheirtermsofservice,17disclosurewaslimitedtospecificareas:● InaFebruary2016blogpost,Twitterdisclosedthat"Sincethemiddleof2015alone,
we’vesuspendedover125,000accountsforthreateningorpromotingterroristacts,"18andinafollow-uppostsixmonthslateritannouncedthat"wehavesuspendedanadditional235,000accounts."19InMarch2017(afterthecutoffdateforinformationevaluatedinthe2017Index)Twitterbeganincludingtermsofservice-relatedtakedownsinitstransparencyreportsforrequeststhatoriginatedfromgovernments
17 Seehttps://rankingdigitalrights.org/index2017/indicators/#F418 https://blog.twitter.com/official/en_us/a/2016/combating-violent-extremism.html19 https://blog.twitter.com/official/en_us/a/2016/an-update-on-our-efforts-to-combat-violent-extremism.html
12
andconcernedextremistcontent.InSeptember2017,Twitterexpandedthiscategorytoincludethreeadditionaltermsofserviceviolationcategories,suchasabusivebehavior,copyright,andtrademark.However,thisdataislimitedtorequeststhatoriginatedfromgovernmentsanddoesnotincluderequestsfromnon-governmentalparties.AsTwitterhasrecentlyannouncedupcomingchangestoitsrules,20specificallythoserelatingtoabuse,itisimportantthatitcontinuetopublishthisdata,andbroadenthescopetoincludenon-governmentrequests,tohelpusersandadvocatesevaluatethenewrules’impact.
● Microsoftpublishedtermsofserviceenforcementdatafor"revengeporn"content(this
contentisillegalinsomejurisdictionsbutnotall),andthecompanyspecificallystatesthattheyremovereportedlinkstophotosandvideosfromsearchresultsonBing"...whenwearenotifiedbyanidentifiablevictim"(whichwouldindicatethatthesearenon-governmentrequests).However,itdidnotpublishdatarelatingtoothertypesofcontentoractivitiesinthecompany'stermsofservice.
● InaSeptember2016blogpost,21YouTubedisclosedthatithadremoved92million
videosforviolatingitstermsofserviceandthat1%ofthevideosremovedwereforhatespeechandterroristcontent.However,itdidnotprovideexactnumbers,andthisdataisnotreportedinanongoingmanner.
Thoughdisclosureonthisindicatorisincrediblylow,itstillmarksanimprovementcomparedtothe2015Index,22inwhichnocompanyevaluateddisclosedanydataaboutcontentthatwasrestrictedfromenforcingtheirtermsofservice.
2.Impactassessment
Sinceitsinceptionin2015,theRDRCorporateAccountabilityIndexhasconsistentlyfoundthatwhilesomeoftheworld’smostpowerfulinternetplatformspubliclydisclosethattheycarryoutimpactassessmentsonhowcompliancewithgovernmentlawsandpoliciesmayaffectthefreedomofexpressionrightsofusers,companiesdiscloselittleaboutassessingtheriskstofreedomofexpressionposedbytheenforcementoftheirownpolicies.Significantchallengestofreedomofexpressionandprivacycanarisewhenacompanydecidestointroduceanewfeature,launchanewservice,orenteranewmarket.Oneindicationthata 20 https://blog.twitter.com/official/en_us/topics/company/2017/safetycalendar.html21 https://youtube.googleblog.com/2016/09/why-flagging-matters.html22 https://rankingdigitalrights.org/index2015/
13
companyisconsideringthepotentialhumanrightsimplicationsofitspoliciesandservicesiswhetheritdisclosesthatitconductshumanrightsimpactassessments(HRIAs).TheUNGuidingPrinciplesonBusinessandHumanRights,whicharticulatebusinesses’responsibilitytorespecthumanrights,specificallyspellsoutcompanies’obligationtoassessactualandpotentialhumanrightsimpactsandtoactuponthefindings.23Humanrightsimpactassessmentsprovidecompanieswithameanstoidentifyareasofconcerninordertomitigateorpreventpotentialinfringementsonhumanrights,ortoprovideremedyforviolationsthatmayhavealreadyoccurred.24Manyoftheissuesattheintersectionofhumanrightsandtechnologymakingheadlinestodaycanbetracedbacktoacompany’sfailuretoanticipatethenegativeimplicationsofitsbusinessdecisions.Forexample,aProPublicainvestigationofFacebook’sinternalcontentmoderationpoliciesrevealedconfusingrulesthatprotectedcategoriessuchas“whitemen”butnot“blackchildren.”Anotherrule,whichthecompanysaidwasnolongerineffect,prohibitedcontentsupporting“violencetoresistoccupationofaninternationallyrecognizedstate.”25ProPublicareportedseveralinstancesofjournalistsandactivistsinPalestine,Kashmir,Crimea,andWesternSahara,whohadtheircontentoraccountsrestrictedasaresultofthispolicy.Suchproblemsraisequestionsaboutwhetherthecompanycarriedoutanysortofimpactassessmentbeforeenactingtheserulesandrelatedenforcementprocesses.Indeed,Facebook’sscoreonIndexindicatorG4,whichexaminescompanydisclosuresonimpactassessment,showsnodisclosureaboutimpactassessmentontermsofservicepolicyformulationorenforcement.26Whencraftingrulesaboutwhattypesofspeechisforbiddenfromtheirplatformsandinwhatcontext,companiesshouldconsultwithexternalstakeholdersandcarefullyexamineiftheserulesprotectthosemostlikelytobediscriminatedagainst,anddeterminetheimpactonfreedomofexpressionoftheirenforcement.Itisimportantforcompaniestocarryouthumanrightsimpactassessmentsonaregularbasisandcontinuetoassessthepotentialimpactoftheirproductsandservicesaftertheinitialrollout,andtoconductmeaningfuloutreachtodifferentstakeholdergroupsinordertolearnhowthesetechnologiesaresurfacingvaryingchallengesforarangeofusergroups.Asnotedinthecallforsubmissions,thestandardsandprocessesthataplatformcompanyusestoenforcesitsownrules—suchasthosearticulatedin
23 Principle17,UNGuidingPrinciplesonBusinessandHumanRights,p.17:http://www.ohchr.org/Documents/Publications/GuidingPrinciplesBusinessHR_EN.pdf.24 UNGuidingPrinciplesonBusinessandHumanRights,p.18:http://www.ohchr.org/Documents/Publications/GuidingPrinciplesBusinessHR_EN.pdf.25 https://www.propublica.org/article/facebook-hate-speech-censorship-internal-documents-algorithms 26 Seehttps://rankingdigitalrights.org/index2017/indicators/#G4;https://rankingdigitalrights.org/2017-indicators/#G4andhttps://rankingdigitalrights.org/index2017/companies/facebook/
14
termsofserviceorcommunityguidelinespolicies—havesignificantimplicationsforhumanrights.Recognizingthisrisk,companiesshouldpaycloseattentiontotheimplicationsoftheirtermsofserviceenforcementpolicieswhenconductingHRIAs.Figure5belowillustratestheextenttowhichcompaniesconductthoroughduediligenceacrossthefullrangeofriskstousers’freedomofexpression:
Sevenofthetenplatformcompaniesevaluatedinthe2017Indexdisclosedatleastsomeinformationaboutwhethertheycarryoutregular,comprehensive,andcredibleduediligence,forexample,intheshapeofhumanrightsimpactassessments(HRIA),toexaminehowallaspectsoftheirbusinessaffectfreedomofexpressionandprivacyandtomitigateanyrisksposedbythoseimpacts.27Fourfullydisclosedthataspartoftheirdecision-making,theyconsiderhowlawsaffectfreedomofexpressioninjurisdictionswheretheyoperate.Fivecompaniesreceivedcreditforatleastsomeinformationaboutassessingfreedomofexpressionandprivacyrisksassociatedwithexistingproductsandservices,andsixdisclosedatleastsomeinformationaboutassessingfreedomofexpressionandprivacyrisksassociatedwithanewactivity.However,ofthetenplatformcompanies,onlytwo—YahooandMicrosoft—disclosedanyinformationaboutassessingandmitigatingthefreedomofexpressionandprivacyrisksassociatedwiththeprocessesandmechanismsusedtoenforcetheirtermsofservice. 27https://rankingdigitalrights.org/index2017/indicators/#G4
15
● Yahoo(nowpartofanewcompany,Oath,ownedbyVerizon),receivedfullcreditfor
disclosingthatoneofthecircumstancestriggeringanHRIAisthe“reviewofinternalprocessesormechanismstoenforcepolicies,suchasourtermsofservice,thatmayimpactusers’rightstoprivacyorfreeexpression."28
● Microsoftreceivedpartialcreditfordisclosureabouteffortsthecompanyhastakento
addressthefreedomofexpressionimpactsofhowitenforcesitstermsofusewithregardtoterroristcontentreports.Initsdisclosure,Microsoftalsostatesthatits“decisionsandactionsintheenforcementofthetermsofuseforourservicesdonotchangebasedonwhetherthereferralismadebyagovernmentoranyothernon-governmententityorperson."29However,thisisonlyoneexampleandisnotaclearcommitmentthatthiskindofassessmentoccursforallitsproductsandservicesonanongoingbasis.
Withtheexceptionsnotedabove,socialandsearchplatformcompaniesarenotdisclosingiftheyconductriskassessmentsofthefreedomofexpressionrisksassociatedwithenforcingtheirtermsofservice.Muchofcompanydisclosuretendstofocusontheirprocessesforrespondingtogovernmentrequeststocensorcontent—companiesarefarlesstransparentwhenitcomestotheimpactthattheirownrules,andprivatepartyrequeststocensorcontent,canhaveonfreedomofexpression.Ifcompaniesarealreadyconductingsuchassessmentsintheseareas,theyshouldbettercommunicatethemtotheirusers.Internetplatformsmustbemakinginformeddecisionsabouttheircontentpolicies—unlesstheyconsultwithaffectedstakeholdersandthoroughlyconsiderandtakeeffortstomitigatehumanrightsharms,thesepoliciesmayhavesignificantnegativeconsequences,whichinmanycasescouldhaveotherwisebeenanticipated.
3.GrievanceandRemedy
Companiesshouldprovidemeaningfulremedywhentheybecomeawareofaninstanceinwhichtheirbusinessoperationsmayhaveresultedintheinfringementofusers’rights,yetfewcompaniesdisclosemuchinformationabouttheirremedymechanisms,iftheyhavethematall.TheUNGuidingPrinciplesonBusinessandHumanRightsstatethatwhenacompanyidentifiesasituationinwhichithascausedorcontributedtoadverseimpactsonhumanrights,”[...]itsresponsibilitytorespecthumanrightsrequiresactiveengagementinremediation,byitselforin 28 https://yahoobhrp.tumblr.com/post/75507678786/human-rights-impact-assessments-yahoo-has29 MicrosoftSalientHumanRightsIssueReporthttp://download.microsoft.com/download/0/0/6/00604579-134B-4D0E-97C3-D525DFB7890A/Microsoft_Salient_Human_Rights_Issues_Report-FY17.pdf
16
cooperationwithotheractors.”30Forinternetplatforms,remedymechanismsmayincludedisclosingaclearprocessbywhichuserscanchallengetermsofserviceenforcementactionsthecompanyhastakenagainstthemandappealforreinstatementoftheircontentoraccount.Companiesshouldalsomakesurethattheirremedymechanismsarebroadenoughtocoverarangeofcomplaintsthatusersmaysubmit.Offlinepowerstructuresareoftenreplicatedonline,andinmaintainingremedymechanisms,platformsshouldthereforeseektoensurethatmarginalizedvoicesareheard.However,thisisnotcurrentlythecase.Forexample,severalRohingyaactivistshavereportedthattheirFacebookaccountsand/orcontenthavebeenrepeatedlycensored.31Thecontentatissuerangedfromnewsaboutmilitaryatrocities,newsaboutmilitaryactioninRakhinestate,andevenapoemaboutrefugeesfleeingmilitaryviolence.“Ihavedeactivatedmyaccountinfrustration,”oneindividualtoldtheDailyBeast.TheseactivistsrelyonplatformslikeFacebooktohelpthemspreadawarenessandnewsofaconflictoftenneglectedbymainstreammedia,butFacebook’srepeateddeletingoftheircontentandthreatstoremovetheiraccountssilencesvoicesspeakingoutaboutanddocumentingatrocities.Incasessuchasthis,thelackofcleargrievanceandremedymechanismcanexacerbatetheoriginalfreedomofexpressionconcern,escalatingthesituationfromaninstanceofcensorshiptoexilefromtheplatformaltogether(whetherself-imposedorasaresultofthecompanydeactivatingtheuser’saccount).Facebookdoesallowuserstoappealsometypesofdecisionsbutnotall(userscannotappealdecisionstoremovalindividualposts,forexample),32anddescriptionsoftheappealsprocessrevealthatitisnotalwaysstraightforwardoraddressedtotheuser’sactualcomplaint.33Atpresent,grievanceandremedymechanismscompaniesofferaretotallyinadequatetomatchtheenormousinfluencetheseplatformswieldoverfreedomofexpression.Figure6onthenextpageillustratestheperformanceoftheteninternetplatformsevaluatedonindicatorG6,whichexaminescompanydisclosuresabouttheirgrievanceandremedymechanisms.
30 UNGuidingPrinciples,p.24:http://www.ohchr.org/Documents/Publications/GuidingPrinciplesBusinessHR_EN.pdf31https://www.thedailybeast.com/exclusive-rohingya-activists-say-facebook-silences-them32 https://onlinecensorship.org/resources/how-to-appeal33 https://www.eff.org/deeplinks/2015/03/facebook-has-clarified-its-policies-how-about-fixing-them
17
Aspressurefromgovernmentsaroundtheworldtopolicecontentcontinuestorise,havingrobustgrievanceandremedymechanismsisessentialtomitigatingharmstofreedomofexpressionandcorrectingerrors.Socialandsearchplatformsatpresentarenotputtingsufficienteffortintogrievanceandremedy,asthe2017Indexdatashows.Notably:● Ofthetenplatformcompaniesevaluated,Kakaohadthehighestscore,with58%.While
thisdisclosurewaslargelyduetorequirementsunderSouthKoreanlaw,Kakaowentbeyondthelegalrequirementforcompliancebyalsoprovidinguserswithanappealsmechanismwhencontentisremovedinresponsetodefamationclaims.Thisthereforeisnotonlyanexampleofregulationplayingapositiverole,butalsoofacompanygoingaboveandbeyondtheminimumlegalrequirement.
● Onlytwoplatformcompanies—KakaoandMail.Ru—fullydisclosedacrossallservices
thattheirgrievanceandremedymechanismsincludecomplaintsrelatedtofreedomofexpression.YouTubealsoreceivedfullcreditonthiselement,thoughGoogleSearchreceivedpartialcredit(G6,Element2,seeappendix).
18
● Twitterwastheonlyplatformcompanytoreceiveanycreditfordisclosingthenumberofcomplaintsitreceivesrelatedtofreedomofexpression.However,itreceivedpartialcredit,asthecopyrightnoticessectionofitsTransparencyReportonlyincludesnumbersofDMCAtakedownsandnototherformsofcomplaintsinvolvingfreedomofexpression.34
Havingadequateremedymechanismsinplacetoaddressfreedomofexpressionharmsisespeciallyimportantinlightofincreasinggovernmentpressureforplatformstoregulateextremistcontent.Thishascomeintheformofnewlaws,suchasGermany’sNetzDGlaw,mentionedearlierinthissubmission,aswellasfromgovernmentspubliclycallingonplatformstostepupeffortstoremoveextremistcontent.Companyattemptstoensurecomplianceappearstobecausingthemtoerronthesideofcensorship—theconsequencesofwhichcanbesevereasdemonstratedbypreviouslycitedexamplessuchasthesilencingofRohingyavoicesinMyanmarorSyrianoppositionactivists.Remedymechanismscanmitigatetheharm—forexample,sothatthousandsofvideosdepictingviolenceinSyriacanberestoredandpreservedforpotentialuseinfuturewarcrimeprosecutions.35Effectivegrievancereportingmechanismsmayalsohelptoinformgovernmentsandcompaniesabouttheconsequencesofcertaintypesofpolicingmechanisms,inadditiontohelpingcompaniesandpolicymakersdevelopalternativepoliciesandpracticesthatresultinlesscollateralcensorship.
Conclusion
Althoughtheinformationthatinternetplatformscurrentlydiscloseshedssomelightonthestateofcontentregulationinthedigitalage,overall,usersandadvocatesarestillleftinthedark.Thelackofcompanytransparencyisespeciallyconcerninginapoliticalcontextwheregovernmentsandotherstakeholdersconcernedabouthatespeech,extremism,andothermaliciousbehavioronlinearepushingforgreatercontentrestrictions.Thepastyearhasseenawidevarietyofeffortstopushcompaniestomorevigorouslypolicetheirplatforms,fromproposalsforautomatedsystemstodetectandfiltercopyright,36andremoveextremistcontent37tocombatinghateandabuse.38Theseproposalshavedifferentscopesandaims,butuntilcompaniesaremoretransparentabouttheirownpolicies,andhowtheyrespondtothirdpartyrequests,itisunclearwhatfurtherimpactthenewlyproposedmeasuresmayhaveonusers’freedomofexpressionrights.Itisdifficulttogaugethepotentialimpacttofreedomof 34 https://transparency.twitter.com/copyright-notices/2015/jul-dec35 http://www.bbc.com/news/technology-4102323436 https://www.eff.org/deeplinks/2017/10/digital-rights-groups-demand-deletion-unlawful-filtering-mandate-proposed-eu37 https://cdt.org/blog/pressuring-platforms-to-censor-content-is-wrong-approach-to-combatting-terrorism/38 https://www.theverge.com/2017/10/14/16410348/twitter-boycott-new-rules-enforcement-aggressive-stance-harassment-jack-dorsey
19
expressionwhenlittledataaboutcontentrestrictionsisavailable,whencompaniesthemselvesdonotappeartobeconductingcomprehensiveimpactassessments,andwhenuserswhosevoicesaresilenceddonothaveadequategrievanceandremedymechanismsthroughwhichtolodgecomplaintsandrestoretheircontent.Meanwhile,freedomofexpressionadvocateswarnthatthetrendtowardsprivatizingcensorship-leavingtherule-makingandenforcementtocompanieswithminimalgovernmentorcourtinvolvement-raisessignificantfreedomofexpressionrisks,andthattheseeffortscomewithsignificantnegativeconsequences.39RDRagreeswiththosewhoarguethatlegislationthatincreasesintermediaryliabilitywillnotadvancefreedomofexpressionbutratherwouldcreateadditionalrisksforviolatingit.Ontheotherhand,therehasbeenrelativelylittlediscussionamongstakeholdersaboutlawsorincentivesthatwouldincreasetransparencyabouthowexpressionisgovernedonlinebycompaniesandgovernments.Measuresrequiringorincentivizinginternetplatformstoprovidegreatertransparencytousersandthepublic,toconductriskassessmentsonpotentialthreatstheirservicesandpoliciesmayhavetofreedomofexpression,andofferingrobustremedymechanisms,couldpotentiallyimproveaccountabilityamongallactorsandwouldalsohelptoinformbettersolutionstotherealproblemofmaliciouscontentandbehavioronline.Corporatetransparency,however,mustbematchedbycommitmentfromgovernmentstodisclosecomprehensivedataaboutthevolumeandnatureofrequestsbeingmadetocompanies.Governmentsshouldsimilarlyconducthumanrightsimpactassessmentstoidentifypotentialadverseimpactsthatmaybecausedbytheenforcementoflawstargetingcontentoninternetplatforms.Finally,governmentsshouldensurethateffectivelegalandothertypesofremediesareavailabletothosewhosefreedomofexpressionrightsareinfringedwhencompaniesattempttocomplywithlawsandotherrequirementstopolicecontent.
39 https://cdt.org/blog/pressuring-platforms-to-censor-content-is-wrong-approach-to-combatting-terrorism/
20
Recommendations
Forcompanies:
● Improvetransparencyandaccountabilityaboutalltypesofthird-partyrequeststorestrictcontentoruseraccounts—governmentrequestsaswellasrequestsbyprivateindividualsandorganizations.Tothemaximumextentpossibleunderthelaw,companiesshouldpublishcomprehensiveinformation(includingtransparencyreports)relatedtothefollowingtypesofthird-partyrequests:
○ Processforrespondingtoalltypesofthird-partyrequeststorestrictcontent,
access,orservice;
○ Dataaboutgovernmentrequeststorestrictcontent,access,orservice;
○ Dataaboutprivaterequestsforcontentrestriction.
● Publishdataonaregularbasisaboutthevolumeandnatureofcontentremovalsandaccountrestrictionsthatthecompanymakestoenforceitstermsofservice.
● Conductregularassessmentstodeterminetheimpactofthecompany'sproducts,
services,andbusinessoperationsonusers'freedomofexpressionandprivacy.Companiesshouldcarryouthumanrightsimpactassessmentsonaregularbasisandcontinuetoassessthepotentialimpactoftheirproductsandservicesaftertheinitialrollout.CompaniesshouldalsopaycloseattentiontotheimplicationsoftheirtermsofserviceenforcementpolicieswhenconductingHRIAs.
● Conductmeaningfuloutreachtodifferentstakeholdergroupsinordertolearnhow
theirservicesmaybenegativelyimpactingusersindifferentcommunities,andidentifywaystomitigatetheseharms.
● Establisheffectivegrievanceandremedymechanisms,andclearlyindicatethatthese
mechanismscanbeusedtoraiseconcernsrelatedtopotentialoractualviolationsoffreedomofexpressionandprivacy.
21
Forgovernments:
● Insteadofpassinglawsthatwouldincreaseintermediaryliabilityoninternetplatforms,governmentsshouldperhapsconsiderregulationrequiringgreatertransparencyfrominternetplatforms.Thiswouldnotonlyhelptoadvancefreedomofexpressionontheseplatforms,butcouldalsohelptoaddressongoingconcernsoftheiruseinpotentialmisinformationandmanipulationofpublicopinion.40
● Considerthepotentialforregulationrequiringcompaniestoconductriskassessments
onpotentialthreatstheirservicesandpoliciesmayhavetofreedomofexpression,andofferrobustremedymechanisms.
● Ensurethatlawsandregulationsallowcompaniestobetransparentandaccountable
withusersabouthowtheyreceiveandhandlegovernmentrequeststorestrictcontent.
● Conducthumanrightsimpactassessmentsonallproposedandexistinglawsrelatedtotheregulationofonlineplatformstoidentifyandmitigatepotentialviolationsofusers’rightstofreedomofexpression.
● Workwithcompaniestoestablisheffectivelegalandothertypesofgrievanceand
remedymechanismsforpeoplewhoserighttofreedomofexpressionhasbeeninfringedinthecourseofpolicingplatformsformaliciouscontent.
40 https://www.nytimes.com/2017/10/29/business/facebook-misinformation-abroad.html
22
Appendix:Fulltextofrelevant2017CorporateAccountabilityIndexindicatorsThefullmethodology,includingdefinitionsandresearchguidance,usedforthe2017Indexcanbedownloadedat:https://rankingdigitalrights.org/wp-content/uploads/2016/09/2017Indexmethodology.pdfRawresearchdataforallindicators,alongwiththefull2017Indexreportcanbedownloadedat:https://rankingdigitalrights.org/index2017/download/Fulltextofindicatorsdiscussedinthissubmissionisduplicatedbelow,pluslinkstoindicatorresearchguidanceanddefinitionsaswellaslinkstotherelevant2017Indexresultsforeachindicator.IndicatorsF3-F7assesstransparencyaboutcompanyactionsaffectingfreedomofexpression.IndicatorG4assessesdisclosureofimpactassessment.IndicatorG6assessesdisclosureofgrievanceandremedymechanisms.F3.Processfortermsofserviceenforcement https://rankingdigitalrights.org/2017-indicators/#F3Thecompanyshouldclearlydisclosethecircumstancesunderwhichitmayrestrictcontentoruseraccounts.Results:https://rankingdigitalrights.org/index2017/indicators/#F3 Elements:
1. Doesthecompanyclearlydisclosewhattypesofcontentoractivitiesitdoesnotpermit?
2. Doesthecompanyclearlydisclosewhyitmayrestrictauser’saccount?
3. Doesthecompanyclearlydiscloseinformationabouttheprocessesitusestoidentifycontentoraccountsthatviolatethecompany’srules?
4. Doesthecompanyclearlydisclosewhetheranygovernmentauthoritiesreceivepriority
considerationwhenflaggingcontenttoberestrictedforviolatingthecompany’srules?
5. Doesthecompanyclearlydisclosewhetheranyprivateentitiesreceivepriorityconsiderationwhenflaggingcontenttoberestrictedforviolatingthecompany’srules?
6. Doesthecompanyclearlydiscloseitsprocessforenforcingitsrules?
7. Doesthecompanyprovideclearexamplestohelptheuserunderstandwhattherulesareand
howtheyareenforced?
23
F4.Dataabouttermsofserviceenforcement https://rankingdigitalrights.org/2017-indicators/#F4Thecompanyshouldclearlydiscloseandregularlypublishdataaboutthevolumeandnatureofactionstakentorestrictcontentoraccountsthatviolatethecompany’srules.Results:https://rankingdigitalrights.org/index2017/indicators/#F4 Elements:
1. Doesthecompanyclearlydisclosedataaboutthevolumeandnatureofcontentandaccountsrestrictedforviolatingthecompany’srules?
2. Doesthecompanypublishthisdataatleastonceayear?
Canthedatapublishedbythecompanybeexportedasastructureddatafile?
F5.Processforrespondingtothird-partyrequestsforcontentoraccountrestriction https://rankingdigitalrights.org/2017-indicators/#F5Thecompanyshouldclearlydiscloseitsprocessforrespondingtogovernmentrequests(includingjudicialorders)andprivaterequeststoremove,filter,orrestrictcontentoraccounts.Results:https://rankingdigitalrights.org/index2017/indicators/#F5 Elements:
1. Doesthecompanyclearlydiscloseitsprocessforrespondingtonon-judicialgovernmentrequests?
2. Doesthecompanyclearlydiscloseitsprocessforrespondingtocourtorders?
3. Doesthecompanyclearlydiscloseitsprocessforrespondingtogovernmentrequestsfrom
foreignjurisdictions?
4. Doesthecompanyclearlydiscloseitsprocessforrespondingtoprivaterequests?
5. Dothecompany’sexplanationsclearlydisclosethelegalbasisunderwhichitmaycomplywithgovernmentrequests?
6. Dothecompany’sexplanationsclearlydisclosethebasisunderwhichitmaycomplywith
privaterequests?
24
7. Doesthecompanyclearlydisclosethatitcarriesoutduediligenceongovernmentrequestsbeforedecidinghowtorespond?
8. Doesthecompanyclearlydisclosethatitcarriesoutduediligenceonprivaterequestsbefore
decidinghowtorespond?
9. Doesthecompanycommittopushbackoninappropriateoroverbroadrequestsmadebygovernments?
10. Doesthecompanycommittopushbackoninappropriateoroverbroadprivaterequests?
11. Doesthecompanyprovideclearguidanceorexamplesofimplementationofitsprocessof
respondingtogovernmentrequests?
12. Doesthecompanyprovideclearguidanceorexamplesofimplementationofitsprocessofrespondingtoprivaterequests?
F6.Dataaboutgovernmentrequestsforcontentoraccountrestriction https://rankingdigitalrights.org/2017-indicators/#F6Thecompanyshouldregularlypublishdataaboutgovernmentrequests(includingjudicialorders)toremove,filter,orrestrictcontentoraccounts.Results:https://rankingdigitalrights.org/index2017/indicators/#F6
Elements:
1. Doesthecompanybreakoutthenumberofrequestsitreceivesbycountry?
2. Doesthecompanylistthenumberofaccountsaffected?
3. DoesthecompanylistthenumberofpiecesofcontentorURLsaffected?
4. Doesthecompanylistthetypesofsubjectmatterassociatedwiththerequestsitreceives?
5. Doesthecompanylistthenumberofrequeststhatcomefromdifferentlegalauthorities?
6. Doesthecompanylistthenumberofrequestswithwhichitcomplied?
7. Doesthecompanypublishtheoriginalrequestsordisclosethatitprovidescopiestoapublicthird-partyarchive?
8. Doesthecompanyreportsthisdataatleastonceayear?
25
9. Canthedatabeexportedasastructureddatafile?
F7.Dataaboutprivaterequestsforcontentoraccountrestrictionhttps://rankingdigitalrights.org/2017-indicators/#F7Thecompanyshouldregularlypublishdataaboutprivaterequeststoremove,filter,orrestrictaccesstocontentoraccounts.Results:https://rankingdigitalrights.org/index2017/indicators/#F7 Elements:
1. Doesthecompanybreakoutthenumberofrequestsitreceivesbycountry?
2. Doesthecompanylistthenumberofaccountsaffected?
3. DoesthecompanylistthenumberofpiecesofcontentorURLsaffected?
4. Doesthecompanylistthereasonsforremovalassociatedwiththerequestsitreceives?
5. Doesthecompanydescribethetypesofpartiesfromwhichitreceivesrequests?
6. Doesthecompanylistthenumberofrequestsitcompliedwith?
7. Doesthecompanypublishtheoriginalrequestsordisclosethatitprovidescopiestoapublicthird-partyarchive?
8. Doesthecompanyreportthisdataatleastonceayear?
9. Canthedatabeexportedasastructureddatafile?
10. Doesthecompanyclearlydisclosethatitsreportingcoversalltypesofprivaterequeststhatit
receives? G4.Impactassessmenthttps://rankingdigitalrights.org/2017-indicators/#G4
26
Thecompanyshouldconductregular,comprehensive,andcredibleduediligence,suchashumanrightsimpactassessments,toidentifyhowallaspectsofitsbusinessaffectfreedomofexpressionandprivacyandtomitigateanyrisksposedbythoseimpacts.Results:https://rankingdigitalrights.org/index2017/indicators/#G4
Elements:
1. Aspartofitsdecision-making,doesthecompanyconsiderhowlawsaffectfreedomofexpressionandprivacyinjurisdictionswhereitoperates?
2. Doesthecompanyregularlyassessfreedomofexpressionandprivacyrisksassociatedwith
existingproductsandservices?
3. Doesthecompanyassessfreedomofexpressionandprivacyrisksassociatedwithanewactivity,includingthelaunchand/oracquisitionofnewproducts,services,orcompaniesorentryintonewmarkets?
4. Doesthecompanyassessfreedomofexpressionandprivacyrisksassociatedwiththeprocesses
andmechanismsusedtoenforceitstermsofservice?
5. Doesthecompanyconductadditionalevaluationwhereverthecompany’sriskassessmentsidentifyconcerns?
6. Doseniorexecutivesand/ormembersofthecompany’sboardofdirectorsreviewandconsider
theresultsofassessmentsandduediligenceintheirdecision-making?
7. Doesthecompanyconductassessmentsonaregularschedule?
8. Arethecompany’sassessmentsassuredbyanexternalthirdparty?
9. Istheexternalthirdpartythatassurestheassessmentaccreditedtoarelevantandreputablehumanrightsstandardbyacredibleorganization?
G6.Remedyhttps://rankingdigitalrights.org/2017-indicators/#G6
27
Thecompanyshouldhavegrievanceandremedymechanismstoaddressusers’freedomofexpressionandprivacyconcerns.Results:https://rankingdigitalrights.org/index2017/indicators/#G6
Elements:
1. Doesthecompanyclearlydiscloseitsprocessesforreceivingcomplaints?
2. Doesthecompanyclearlydisclosethatitsprocessincludescomplaintsrelatedtofreedomofexpressionandprivacy?
3. Doesthecompanyclearlydiscloseitsprocessforrespondingtocomplaints?
4. Doesthecompanyreportonthenumberofcomplaintsreceivedrelatedtofreedomof
expressionandprivacy?
5. Doesthecompanyprovideclearevidencethatitisrespondingtocomplaints?