2024年AI時(shí)代來臨：青少年與家長在家校環(huán)境中對(duì)生成式AI的采納與態(tài)度研究

2024

TheDawnoftheAIEra:

Teens,Parents,andtheAdoptionofGenerativeAIatHomeandSchool

COMMONSENSEISGRATEFULFORTHEGENEROUSSUPPORTANDUNDERWRITINGTHATFUNDEDTHISRESEARCHREPORT

CarnegieCorporationofNewYorkJenniferCaldwellandJohnH.N.Fisher

CraigNewmarkPhilanthropiesPatrickJ.McGovernFoundationSiegelFamilyEndowment

2024

TheDawnoftheAIEra:

Teens,Parents,andtheAdoptionofGenerativeAIatHomeandSchool

Credits

Acknowledgments:ThisprojectrepresentstheworkofmanypeoplefromCommonSenseMediaandbeyond.

Below,wedetaileachpartoftheprojectandthecontributors.

Projectadvisors:NaomiBaron,CelesteKidd,JasmineMcNealy,andLeeRainie

SurveydevelopmentAmandaLenhartandMaryMadden,withsupportfromAngelaCalvinandAlexaHasse.

anddesign:ThankstoJulieValdiviaforherassistancewiththesurveytranslation.

Datacollection:IpsosU.S.PublicAffairs

DataanalysisandAngelaCalvin,AlexaHasse,AmandaLenhart,andMaryMadden

interpretation:

Reportwriting:MaryMadden,AngelaCalvin,andAlexaHasse,withsupportfromAmandaLenhart

Internalreview:EishaBuch,DanielVargasCampos,AminaFazlullah,TracyPizzoFrey,JeffreyKnutson,

MerveLapus,JasmineHoodMiller,EllenPack,andYvetteRenteria

Designer:AliceSeiler

Copyeditors:ChristopherDareandJenniferRobb

Communications:ChristianMedinaBeltz,KirstenOlsenCahill,LisaCohen,NataliaGarcia,

andMarisaNaughton

Thereportauthorsandprojectteamwouldliketoexpresstheirgratitudetoalltheyoungpeopleandtheirparentswhogavetheirtimetotakethesurvey.

Suggestedcitation:Madden,M.,Calvin,A.,Hasse,A.,&Lenhart,A.(2024).ThedawnoftheAIera:Teens,parents,andtheadoptionofgenerativeAIathomeandschool.SanFrancisco,CA:CommonSense.

TableofContents

Introduction

1

KeyFindings

4

Part1:HowTeensandParentsUseGenerativeAITools 17

HowandWhyTeensUseGenerativeAI

20

Part2:HowTeensUseandExperienceAI

inSchoolandforSchool

25

GenerativeAIintheClassroom

26

GenerativeAISchoolRulesandPolicies

31

Part3:GenerativeAIinEducation:

ExploringtheBenefits,Risks,andImpacts

onEquitywithTeensandParents

34

Part4:HowGenerativeAIChangesYouthandParent

PerspectivesontheFuture

43

Conclusion 47

Methodology 49

Introduction

WhenOpenAI'sChatGPTplatformwasreleasedtothepublicinNovember2022,manypunditsandpoliticianswerequicktoheraldthegreatpromiseofgenerativeAI.1Otherpromi-nentvoicesswiftlyhighlightedthepotentialpitfallsofthisnewtechnology:Economistswarnedofarapidlossofjobsduetoautomation,2andnonprofitshighlightedtherisksofunfet-tereddevelopmentwithoutguardrails.Globalleadersexpressedconcernthatsocietywouldfaceexistentialrisks3associatedwithanunprecedenteddelugeofdisinformation,normalizationofbias,andlonger-termthreatstohumanagency.However,whilethesedebatesaboutpotentialimpactsplayedout,educationalinstitutionswerethrustontothefrontlinesinnavigatingthischangeovernight.

ManyoftheimmediateimpactsofChatGPTandothergenera-tiveAItoolsthatfollowedsuitwerefeltincollegesandK–12schools,wherestudentsquicklyfoundtheycouldretrievemonths'worthofresearchandwritinginamatterofseconds.Teacherswereleftscrambling4toidentifywhetherstudentshadsubmittedAI-generatedcontentasoriginalwork,andstu-dentspostedscoresoftutorialsonsocialmediadescribingwaystocircumventtoolsdesignedtodetectAI-drivenplagiarism.

Morerecently,schoolshavefacedemergingchallengesthatincludegenerativeAIcontentbeingusedtoimpersonatestaff5andtheamplificationofvariousformsofharassment,includingtheproductionofdeepfakeimageryofstudents.6

Atthesametime,schoolshavealsobeenatestinggroundforexperimentationwiththepositivepotentialforgenerativeAItoaugmentandsupportstudentlearning.Recentstudieshavesuggestedthatinitialfearsaboutwidespreadincreasesincheatingmayhavebeenoverblown,7andnationalsurveysthatreportstudents'attitudesaboutacceptableusesofgenAItoolsinschoolsuggest8thatmostconsiderthemtobeaccept-ablepartnersforresearchandeditorialsupport,butdrawthelineatgeneratingessaysontheirbehalf.

SomeschoolsanddistrictsthatinitiallyrejectedanyuseofChatGPTlatershiftedtoembracethetool9sothatteachersandstudentscouldlearntointegrateitintotheclassroom.Attheuniversitylevel,manyschoolshaveissuedguidancetostudentsandteachers,andhavesuggestedarangeofbeneficialusecases,10suchasbrainstormingnewideas,practicinganewlanguage,developinglessonplans,helpingtowriteandcorrectcomputercode,andidentifyingpatternsinlargedatasets.

1Rotman,D.(2023,March25).ChatGPTisabouttorevolutionizetheeconomy.Weneedtodecidewhatthatlookslike.MITTechnologyReview.

/2023/03/25/1070275/chatgpt-revolutionize-economy-decide-what-looks-like/

2Greenhouse,S.(2023,February8).USexpertswarnAIlikelytokilloffjobs—andwidenwealthinequality.TheGuardian.

/technology/2023/feb/08/ai-chatgpt-jobs-economy-inequality

3Roose,K.(2023,May30).AIposes'riskofextinction,'industryleaderswarn.TheNewYorkTimes.

/2023/05/30/technology/ai-threat-warning.html

4Meckler,L.,&Verma,P.(2022,December28).TeachersareonalertforinevitablecheatingafterreleaseofChatGPT.TheWashingtonPost.

/education/2022/12/28/chatbot-cheating-ai-chatbotgpt-teachers/

5Griffith,K.,&Fenton,J.(2024,April25).Ex-athleticdirectoraccusedofframingprincipalwithAIarrestedatairportwithgun.TheBaltimoreBanner.

/education/k-12-schools/eric-eiswert-ai-audio-baltimore-county-YBJNJAS6OZEE5OQVF5LFOFYN6M/

6Hadero,H.(2023,December2).Teengirlsarebeingvictimizedbydeepfakenudes.Onefamilyispushingformoreprotections.TheSeattleTimes.

/nation-world/nation/as-teen-girls-in-wa-new-jersey-are-being-victimized-by-deepfake-nudes-one-family-

is-pushing-for-protections/

7Spector,C.(2023,October31).WhatdoAIchatbotsreallymeanforstudentsandcheating?StanfordGraduateSchoolofEducation.

/news/what-do-ai-chatbots-really-mean-students-and-cheating

8Sidoti,O.,&Gottfried,J.(2023,November16).About1in5U.S.teenswho'veheardofChatGPThaveuseditforschoolwork.PewResearchCenter.

/short-reads/2023/11/16/about-1-in-5-us-teens-whove-heard-of-chatgpt-have-used-it-for-schoolwork/

9Jones,B.,Touré,M.,&PerezJr.,J.(2023,August23).MoreschoolswantyourkidstouseChatGPT.Really.Politico.

/news/2023/08/23/chatgpt-ai-chatbots-in-classrooms-00111662

10CenterforTeachingInnovation.(2023).Generativeartificialintelligence.CornellUniversity.

/generative-artificial-intelligence

1

THEDAWNOFTHEAIERA:TEENS,PARENTS,ANDTHEADOPTIONOFGENERATIVEAIATHOMEANDSCHOOL?COMMONSENSEMEDIA.ALLRIGHTSRESERVED.

Yetevenwithalltheaddedpowerofdatacollectionandsyn-thesis,thefutureofgenerativeAItoolsisstilluncertain.Manyquestionsremainaboutwhatwillhappentocriticalskills,likedecision-makingandwriting.Otherswonderhowhumanintelligenceandourcapacityforcreativityandlearningmightchangealongsidethesepowerfulagents.Acknowledgingtheseandotheropenquestions,theU.S.DepartmentofEducationreleasedguidancelastfallthatraisedanumberofinquiriesaboutthequalityoftheunderlyingdatausedtotrainlargelanguagemodelsandtheresultingtendencytoproducebiasedresults.11

Andaswithmanycommercialtechnologyapplicationsusedinschools,teachersandadministratorsmustconsidertheprivacyimplicationsofstudentandinstitutionaldatathatmaybecollectedthroughtheirengagementwithgenAIplatforms12

Thisresearchreportisintendedtoprovideadditionaldataandsupportforthosewhoaredevelopingeducational,research,andpolicyinitiativestobetterunderstandandrepresenttheinterestsofmiddleandhighschoolstudentsandtheirparentsorguardiansatatimeofgreatdebateovertheintegrationofartificialintelligencetechnologiesinschools.Drawingonnew,nationallyrepresentativesurveydatafrom1,045teensandtheirparents,thefindingsillustratethegrowingroleofgen-erativeAIplatformsinthelivesoffamiliestoday.

Parents'andtheirteens'awarenessanduseofgenerativeAIdiffer,asdotheirattitudesabouttheperceivedeffectsofthesetechnologiesoneducation,learning,andwork.TheseattitudesvarysignificantlybasedonthelevelofexperiencethatrespondentshavehadwiththeseAItools,afamily'sracialandethnicbackground,andtheirsocioeconomicstatus.Generallyspeaking,youngpeopleandtheircaregiversrecog-nizeamixofpotentialbenefitsandrisksassociatedwitheducationalapplicationsofgenerativeAIplatforms,andmanythinkthesetechnologieswillimpacttheirfutureeducationplansandjobprospects.

Summaryofmethodology

?Thisisanationallyrepresentativesurveythatincludes

responsesfrom1,045generalpopulationU.S.adults(age

18orolder)whoareparentsorguardiansofoneormoreteensage13to18,andresponsesfromoneoftheseteens.All18-year-oldrespondentswerestillinhighschool.

?Paired(dyad)parentandteensurveysincluded

responsesfrom1,045adolescentsage13to18and

oversamplestoobtain250Blackand300Latinoteenrespondents.

?DatawascollectedbyIpsosPublicAffairsonbehalfofCommonSenseMediafromMarchtoMay2024.

?Thesurveywasconductedonline,inEnglishorSpanish.

?Differencesbetweensubgroupsweretestedforstatisticalsignificanceatthelevelofp<.05.

?Totalamountsmaynotsumto100%fromthereportedsubtotalsduetoroundingandnonresponse.

?Foradditionaldetails,pleaseseetheMethodologysectionofthisreport.

11OfficeofEducationalTechnology.(2023,May).Artificialintelligenceandthefutureofteachingandlearning:Insightsandrecommendations.U.S.DepartmentofEducation.

/files/2023/05/ai-future-of-teaching-and-learning-report.pdf

12DivisionofInformationTechnology.(2023,May26).UsinggenerativeAIwhilerespectingprivacy.UniversityofMaryland.

/news/using-generative-ai-while-respecting-privacy

?COMMONSENSEMEDIA.ALLRIGHTSRESERVED.THEDAWNOFTHEAIERA:TEENS,PARENTS,ANDTHEADOPTIONOFGENERATIVEAIATHOMEANDSCHOOL2

KeyFindings

1.Sevenin10teensage13to18say

theyhaveusedatleastonetypeof

generativeAItool.Searchengines

withAI-generatedresultsandchatbotsareconsiderablymorepopularthan

image-andvideo-generatingtools.

Similartothedramaticgrowthofsocialmedia,mostteenshavebeenquicktoadoptgenerativeAItools.Unlikeotherdigitaltechnologies,theuseofgenerativeAI—includingsearchengineswithAI-generatedresults,chatbots,andtextgeneratorsaswellasimageandvideogenerators—isconsis-tentacrossteensofallages,genders,andracialandethnicbackgrounds.Andwhile70%ofteenssaytheyhaveusedatleastonetypeofgenAItool,teenswithaparentwhohasacollegedegreearesignificantlymorelikelythanthosewithaparentwhohasnotcompletedcollegetohaveusedatleastonetypeofgenerativeAI(74%vs.66%).

Amongteensage13to18,AI-supportedsearchandchatbotsaremorelikelytobeused,whileimageandvideogeneratorsarelesscommon:

?56%ofteenssaytheyhaveusedsearchengineswith

AI-generatedresults,suchasBing,GoogleSGE,orBraveSummarizer.

?51%haveusedchatbots/textgenerators,suchasChatGPT,GoogleGemini,orSnap'sMyAI.

?34%haveusedimagegenerators,suchasDALL-E,PhotoshopAI,orBingImageCreator.

?22%ofteenshaveusedAIvideogenerators,suchasMidjourney,StableDiffusion,orGoogleVideoPoet.

TeensmostoftenuseAI-supportedsearch

PercentofteenswhousethefollowingtypesofgenerativeAItools

SearchenginesNET

20%16%20%56%

Chatbots

11%13%27%51%

Imagegenerators

6%7%21%34%

Videogenerators

7%5%10%22%

DailySeveralTimesPerWeekAFewTimesPerMonthorLess

Note:Totalamountsmaynotsumto100%fromthereportedsubtotalsdueto

roundingandnonresponse.Q:"Howoften,ifever,haveyouusedthefollowing

kindsofgenerativeAIsystems?"Therespondentsincludedinthischartwere1,045youngpeopleage13–18.Source:CommonSenseMediasurveyconductedbyIpsos,March15–April20,2024,with1,045youngpeopleage13–18andtheirparents.

Noteworthydifferencesemergebasedonrace/ethnicityinthetypeofgenerativeAItoolsused.BlackandLatinoteensaresignificantlymorelikelytouseimageandvideogeneratorscomparedtotheirWhitepeers(39%ofBlackteensand40%ofLatinoteensvs.30%ofWhiteteensforimagegenerators,and32%ofBlackteensand31%ofLatinoteensvs.16%ofWhiteteensforvideogenerators).

?COMMONSENSEMEDIA.ALLRIGHTSRESERVED.THEDAWNOFTHEAIERA:TEENS,PARENTS,ANDTHEADOPTIONOFGENERATIVEAIATHOMEANDSCHOOL4

2.TeenssaytheyareusinggenerativeAIforavarietyofpurposes,butusingitforhelpwithhomeworkisthemost

common.

BeyondwhichAIplatformsteenschoosetouse,theyusegen-erativeAIforavarietyoftasksandpurposes.ThemostcommonlyreportedactivitiesareusinggenAIforhomeworkhelp(53%),tostaveoffboredom(42%),andtotranslatesome-thingfromonelanguagetoanother(41%).

Weaskedabout14differentactivitiesinthissurveythatfallintotwocategories:formaloreducationalactivities,andsepa-rately,personalactivities.Formaloreducationalactivitiesincludesummarizingortranslatingcontent,makingimagesorvideos,brainstormingideas,helpingwithhomework,andwritingdocuments,email,andcode.Mostoftheseactivitiesarecarriedoutbyaboutone-thirdanduptoone-halfofallteens,withtheexceptionofwritingcode,whichonlyabout12%ofteenshaveeverengagedinwiththeassistanceofgenAI.

Inthissurvey,wealsoaskedaboutgenAI-supportedpersonalactivitiessuchasstavingoffboredomandloneliness,creatingcontentasajokeortoteasesomeone,creatingnewcontentusingsomeoneelse'svoiceorimage,planningactivities,andseekingpersonalorhealthadvice.Feweryouthturntogen-erativeAIforpersonalactivities.Between12%and19%engageintheseactivitieswiththeassistanceofgenAI.TheoneexceptiontothistrendisusinggenAItorelieveboredom,whichmorethantwoinfiveteensreport.

Inkeepingwithhistoricalpatternsformobilephonesandotherscreenmedia,BlackandLatinoteensaremorelikelytoengageinawidervarietyofactivitieswithgenAIthantheirWhitepeers.13LargersharesofbothBlackandLatinoyouthcomparedwithWhiteyouthsaytheyhaveusedgenerativeAItoolsto:

?Createcontentasajokeortoteaseanotherperson

(24%ofBlackteensand24%ofLatinoteensvs.15%ofWhiteteens)

?Helpthemplananactivity(24%and22%vs.10%)

?Keepthemcompany(26%and18%vs.11%)

?Seekhealth-relatedadvice(16%and22%vs.8%)

?Generatenewcontentfromaperson'svoiceorimage(20%and17%vs.9%)

BlackyoutharemorelikelythanWhiteyouthtousegenera-tiveAIto:

?Gethelpwithhomework(59%ofBlackteensvs.47%ofWhiteteens)

?Getadviceonapersonalissue(25%vs.14%)

?Helpthemwritecodeorcreateanapp(17%vs.7%)

Latinoyouth,bycomparison,aremorelikelythanWhiteandBlackyouthtousegenerativeAItotranslatesomethingfromonelanguagetoanother(53%vs.37%and31%,respectively).LatinoyoutharealsomorelikelythanWhiteyouthtousegenAItowriteadocumentoremail(41%vs.29%)andtocreateanewimageorvideo(43%vs.27%).

13SeePewResearchCenter.(2023,December11).Teens,socialmedia,andtechnology2023.

/internet/2023/12/11/teens-social-media-and-technology-2023/

5

THEDAWNOFTHEAIERA:TEENS,PARENTS,ANDTHEADOPTIONOFGENERATIVEAIATHOMEANDSCHOOL?COMMONSENSEMEDIA.ALLRIGHTSRESERVED.

TeenpurposeofgenerativeAIuse,byrace/ethnicity

AmongteenswhousegenAI,thepercentagewho

haveeverusedittoassisttheminthefollowingways...

TeenswhousegenAI,byrace/ethnicity

Total

White

Black

Latino

Helpwithhomework

53%

47%a

59%b

57%ab

Keepmefrombeingbored

42%

39%a

42%a

48%a

Translatesomethingfromonelanguagetoanother

41%

37%a

31%a

53%b

Brainstormideas

38%

35%a

37%a

43%a

Writeadocumentoremail

35%

29%a

39%ab

41%b

Createanewimageorvideo

33%

27%a

32%ab

43%b

Summarizeorsynthesizeinformation

33%

31%a

30%a

35%a

Createcontentasajokeortoteaseanotherperson

19%

15%a

24%b

24%b

Getadviceonapersonalissue

18%

14%a

25%b

22%ab

Plananactivity

16%

10%a

24%b

22%b

Keepmecompany

15%

11%a

26%b

18%b

Seekhealth-relatedadvice

14%

8%a

16%b

22%b

Generatenewcontentfromaperson'svoiceorimage

12%

9%a

20%b

17%b

Writecodeorcreateanapp

12%

7%a

17%b

12%a

Note:Itemswithdifferentsuperscriptsdiffersignificantlyacrossrowswithineachcategory(p<.05).Q:"HaveyoueverusedgenerativeAItoassistyouwithanyofthefollow-ing?"Therespondentsincludedinthistablewere763youngpeopleage13–18whohaveeverusedgenerativeAI.Source:CommonSenseMediasurveyconductedbyIpsos,March15–April20,2024,with1,045youngpeopleage13–18andtheirparents.

?COMMONSENSEMEDIA.ALLRIGHTSRESERVED.THEDAWNOFTHEAIERA:TEENS,PARENTS,ANDTHEADOPTIONOFGENERATIVEAIATHOMEANDSCHOOL6

3.Twoinfive(40%)teensreportusinggenerativeAIforschoolassignments,withanearlyevensplitbetweenthosewhousegenAIwiththeirteacher's

permissionandthosewhodonot.

While70%ofteensreportusinggenerativeAI,asmallerpro-portionofteensusegenAIspecificallyforschoolwork.

?Twoinfive(40%)teensreporthavingeverusedgenAItohelpwithschoolassignments,while60%saytheyhaveneverusedgenAItohelpwithschoolwork.

?LatinoteensweremorelikelytosaythattheyhaveusedgenAItohelpwithschoolassignmentscomparedtotheirWhitepeers(45%vs.36%).

TeenswhoreportedusinggenerativeAItohelpwithschoolassignmentswerepromptedtothinkaboutthemostrecenttimetheyusedgenAItohelpwithschoolworkandwhetherusinggenAIwaswiththeirteacher'spermissionornot.

?Abouttwoinfive(41%)teenswhousedgenAItohelpwithschoolworkdidthemostrecentassignmentwiththeirteacher'spermission.

?Almostthesameproportionofteens,46%,usedgenAIfortheassignmentwithouttheteacher'spermission.

?About1in10(12%)teenswerenotsurewhetherthey

usedgenAIfortheassignmentwiththeirteacher'sper-missionornot.14

TwoinfiveteensusegenerativeAIforschoolwork,withanearlyevensplitbetweenthoseusingwithandwithoutteacherpermission

PercentofteenswhohaveusedgenAIforschoolassignments,andthosewhodidsowithorwithoutteacher'spermission

41%

Yes,withmy

teacher's

permission

12%

Yes,butunsure

60%

No,Ihavenotused

genAItohelpwith

schoolwork

40%

Yes,Ihaveused

genAItohelp

withschoolwork

ifwithteacher's

permission

46%

Yes,withoutmyteacher'spermission

Note:Totalamountsmaynotsumto100%fromthereportedsubtotalsduetoroundingandnonresponse.Q:"HaveyoueverusedgenerativeAItohelpwithyourschool

assignments?"Ifateenanswered"yes,"theyweregiventhefollowingquestion:"Asareminder,thissurveyisentirelyconfidential.ThinkingaboutthemostrecenttimewhenyouusedgenerativeAItohelpwithaschoolassignment,wasitwithyourteacher'spermissionornot?"Therespondentsincludedinthischartwere452youngpeopleage13–18whohaveeverusedgenerativeAItohelpwithschoolassignments.Source:CommonSenseMediasurveyconductedbyIpsos,March15–April20,2024,with1,045

youngpeopleage13–18andtheirparents.

14Anotherpossibleoptionisthatteensmaynothavebeengivenexplicitpermissionorspecificallydeniedpermissiontousethesetools.Weattemptedtocapturethisusingthe"notsure"option.Thesamplesizesforcertainsubgroupswererelativelysmallforthisquestion,andtherewerenosignificantdifferencesbyage,grade,gender,race/ethnicity,orLGBTQ+identity.

7

THEDAWNOFTHEAIERA:TEENS,PARENTS,ANDTHEADOPTIONOFGENERATIVEAIATHOMEANDSCHOOL?COMMONSENSEMEDIA.ALLRIGHTSRESERVED.

4.Amongteenswhohaveused

generativeAIforschoolassignments,63%saytheyhaveusedchatbotsortextgenerators,whilealittlemorethanhalf(57%)haveusedsearchengineswithAI-generatedresults.

TeenswhohaveusedgenerativeAItohelpwithschoolassign-mentsusevarioustypesofsystemstosupporttheirschoolwork:

?63%haveusedchatbotsortextgenerators.

?57%haveusedsearchengineswithAI-generatedresults.

?23%haveusedimagegeneratorsfortheirschoolwork.

?13%haveusedvideogeneratorsforschool.

?Whiteteensaremorelikelytohaveusedchatbots/textgenerators(68%)comparedtoLatinoteens(53%).

?Latinoteensaremorelikelytohaveusedimagegenera-torsforschoolwork(30%)thanWhiteteens(18%).

?Black(20%)andLatinoteens(18%)areaboutthreetimesmorelikelythanWhiteteens(6%)tosaytheyhaveusedvideogeneratorsforschool.

Chatbots/textgeneratorsarethemostcommonplatformteensuseforschoolwork

PercentofteenswhohaveusedvarioustypesofgenerativeAItohelpwithschoolassignments

Chatbots/textgenerators

63%

SearchengineswithAI-generatedresults

57%

Imagegenerators

23%

Videogenerators

13%

Note:Q:"WhattypesofgenerativeAIhaveyouusedforschoolassignments?"Therespondentsincludedinthischartwere420youngpeopleage13–18whohave

usedgenerativeAItohelpwithschoolassignments.Source:CommonSenseMediasurveyconductedbyIpsos,March15–April20,2024,with1,045youngpeopleage13–18andtheirparents.

5.Mostparentsareinthedarkabouttheirchild'sgenerativeAIuse;just37%ofparentswhoseteenreportedusingatleastonegenAIplatformthoughttheirchildhadalreadyusedgenAI.

Therearesignificantgapsinparents'awarenessoftheirchild'suseofgenerativeAI.AmongparentswhoseteenindicatedusingatleastonegenerativeAIplatform,only37%thoughttheirchildhadusedgenAI.Aboutoneinfour(23%)oftheseparentsbelievetheirchildhadnotusedanyoftheseplat-forms,while39%werenotsurewhethertheirchildhadusedthesetools.

Mostparentsareoutofthelooparoundtheirteen’suseofgenerativeAI

PercentofparentswhoseteenhasusedatleastonegenAIplatformwhoindicateawarenessofthatuse

23%

No

39%

Notsure

37%

Yes

Note:Q:"Toyourknowledge,hasyour13-to18-year-oldchildeverused

generativeAI?"Therespondentsincludedinthischartwere763youngpeopleage13–18whohaveusedatleastonegenerativeAIplatformandtheirparents.Source:CommonSenseMediasurveyconductedbyIpsos,March15–April20,2024,with1,045youngpeopleage13–18andtheirparents.

AlthoughmostparentsareawareofgenAItools—87%haveheardatleast"alittle"aboutgenAI—manyhavenotdiscussedthetopicwiththeirteen.Almosthalf(49%)saytheyhavenottalkedaboutgenAI,while45%havehaddiscussionsaboutthetechnologywiththeirchild.

?COMMONSENSEMEDIA.ALLRIGHTSRESERVED.THEDAWNOFTHEAIERA:TEENS,PARENTS,ANDTHEADOPTIONOFGENERATIVEAIATHOMEANDSCHOOL8

6.About6in10teensreporteither

thattheirschoolhasnorulesforhowgenerativeAIcanbeusedorthat

they'renotsureiftherearerules.

AndmostparentssaythatschoolshavenotcommunicatedwithfamiliesaboutgenAIpolicies.

Basedonteenreports,manyschoolsandteachersmaynothaveclearlycommunicatedaboutorimplementedrulesforgenerativeAI.Andmanyteensstillreportuncertaintyaboutschoolpoliciesandteacherpermissions.

?37%ofteensarenotsurewhethertheirschoolhasgenAIrules.

?35%saytheirschooldoeshaverulesforhowgenAIcanbeused.

?27%reportthattheirschoolhasnorules.

MostteensreportthatteachersdoNOTallowgenerativeAI

PercentofteenswhoindicatethattheirteachershavediscusseduseofgenAIinthefollowingways

MostlydonotallowstudentstousegenAI

42%

TeachershavenotmentionedgenAI

27%

CommunicationandpermissionfromteachersinregardtogenerativeAIforschoolworkvarysignificantly.Andmostparentssaytheirchild'sschoolhasnotcommunicatedwiththemaboutgenerativeAIatall.Yetevenincaseswheretheremaynotbeschoolwiderules,studentsreportthattheirteach-ershaveimplementedarangeofrestrictions:

?42%ofteenssaytheirteachersmostlydonotallowstu-dentstousegenerativeAI.InaseparateCommonSense

2023survey,15amajorityofstudentsage12to18(51%)

feltthatschoolsshouldlimittheuseofgenAIprograms,suchasChatGPT,untilsafeguardsandrulesareinplace.

?27%ofteensinthecurrentsurveysaytheirteachershavenotmentionedgenerativeAI,and23%werenotsurewhethertheirteachersallowgenAIuse.

?7%reportthattheirteachersmostlyallowstudentstousegenerativeAI.

?Notably,Blackteens(13%)aremorelikelytosaytheir

teachersmostlyallowstudentstousegenerativeAI

whencomparedwithWhite(7%)andLatinoteens(7%).

Thelargemajorityofparentsofteens(83%)saythattheirchild'sschoolhasnotcommunicatedwiththemaboutgenera-tiveAI,comparedwith16%ofparentswhoreportthattheschoolhascommunicated.Just10%