1、VOICE CLONE DEEPFAKE ANDREPORTCONSUMER SENTIMENTOCTOBER 2023METHODOLOGY1.INTRODUCTION/32.SCAMS,ATTACK VECTORS,&POSITIVE USES/133.DEEPFAKE AWARENESS&SENTIMENT/234.VOICE CLONE AWARENESS&SENTIMENT /375.COMPARING DEEPFAKES AND VOICE CLONES/526.DEEPFAKE&VOICE CLONE CONCERN BY DOMAIN/607.ADDITIONAL RESOUR

2、CES/71Table of ContentsVoicebot produces the leading online publication,newsletter and podcast focused on the converstional AI and generative AI industries.Thousands of industry leaders,engineers,entrepreneurs,software developers,investors,analysts and company founders look to Voicebot each week for

3、 the latest news,data,analysis and insights defining the trajectory of the next great AI computing platforms.Synthedia is the leading provider of analysis,data,trends,news,and advisory services focused on the rapidly growing generative AI and synthetic media markets.Through a daily email newsletter,

4、weekly news live stream,conferences,reports,community,and strategy consulting,Synthedia helps business leaders navigate the most disruptive technology shift in two decades.The survey was conducted online during July of 2023 and was completed by 2,027 U.S.adults age 18 or older that were representati

5、ve of U.S.Census demographic averages.Because we reached only online adults which represent 95%of the population,some totals are adjusted downward to provide usage numbers relevant to the entire adult population.Pindrop was built to silence fraudsters.AI fraud is on the rise,making cutting-edge frau

6、d detection software your key to stopping deepfake attacks.Pindrop is dedicated to continuously innovating solutions to detect and mitigate the risks posed by synthetic audiofortifying trust and integrity between you and your customers.INTRODUCTION VOICEBOT.AI-All Rights Reserved 2023A Letter From P

7、indropcombination of social engineering and technology in attempts to access someones bank account,medical history,or other private information.Today,it is the same,but voice clones and deepfakes are the latest tools in the fraud technology arms race.Pindrop is taking action and providing detection

8、tools to protect companies and individuals from harm.In fact,we have been working on this problem for a decade and have 11 deepfake detection patents,many of which have been converted into products.We collaborated with Voicebot.ai to shed light on consumer experience and perception around deepfake a

9、nd voice clone technology.I suspect you will find that deepfake awareness by over half of U.S.adults is higher than expected.Voice clone awareness is even higher at 64%.The proliferation of the technology has moved quickly,and consumers are rightly concerned.Thankfully,there are solutions for compan

10、ies that want to proactively protect their customers.Vijay Balasubramaniyan,CEO and co-founder of PindropEach new technology arrives with applications for good and for bad.While many fans of Star Wars,Americas Got Talent,online games,and countless humorous YouTube videos have enjoyed the emergence o

11、f voice clones and deepfakes,there is also a rising use of these technologies in fraud,disinformation,and harm to personal reputations.The technologies are so good,and we are inundated with so much digital media that people now struggle to differentiate the real from the fake.A Pindrop study showed

12、that people can identify a deepfake with just 57%accuracy.That is only 7%better than a coin toss.The solution is not to stop the technology and throw out the good in order to eliminate the bad.The solution is to create tools that can detect deepfakes and voice clones,and protect users from the risks

13、 of misuse.Pindrop Security was founded on using AI technology to reduce risk by distinguishing reality from the illusory.Initially,it was humans employing a PAGE 3The Rise of DeepfakesINTRODUCTIONPAGE 4 VOICEBOT.AI-All Rights Reserved 2023Most of the news coverage and analysis focuses on the techno

14、logy and events and little is available about consumer perception and experience with deepfakes.Voicebot partnered with Pindrop Security in an effort to fill this information gap about deepfakes overall and for the subcategory of voice clones.The findings may surprise you.We did not anticipate the l

15、evel of everyday consumer awareness or the split of positive and negative sentiment,especially in light of the level of concern expressed.Nor did we expect so many consumers to believe that banks,insurance,healthcare,news,and social media organizations have already taken meaningful steps to protect

16、them.This finding suggests that there may be a mismatch between expectations and reality.That sentiment surely seems on topic.Deepfakes are one of the most controversial solutions arising from the advancement of AI technology.The technology is used by Hollywood to de-age actors and recreate actors l

17、ikenesses after death,by comedians in social media,by advertisers,by gamers,and among friends for fun and entertainment purposes.However,it is also employed by criminals as a tool to commit fraud,by disgruntled people looking to harm someones reputation,and by nation states and political actors to s

18、pread disinformation.Whether you view deepfake technology as net positive or net negative,its unreality is a reality of our time.Voicebot.ai and Synthedia have published over 100 articles on deepfake technology since 2019.Most of those in the past year.The technology is on the rise,improving in qual

19、ity,and becoming more accessible.Filling an Information GapINTRODUCTIONPAGE 5 VOICEBOT.AI-All Rights Reserved 2023Deepfakes Exist in Multiple FormatsVoicebot categorizes deepfake technology into four categories:audio,video,text,and images.The core technologies are generally related but distinct.For

20、this report,we focus on the overall deepfake market and the subcategory of voice clones.The various technologies can be applied alone or combined and may be static,recorded,or delivered in real time.DefinitionsDeepfakes are AI-manipulated digital media in text,image,audio,or video formats that repli

21、cate something that is real or alter key characteristics of something that is real which may change how the media is interpreted.The word is a portmanteau combining the technical term deep learning with fake.Voice Clones are AI-manipulated audio media that replicate the voice of a specific person.It

22、 is a type of synthetic voice and a subcategory of deepfakes.Deepfake Technology CategoriesAudioVoicesMusicSoundsVideoDynamicPeopleObjectsPlacesConceptsPeopleObjectsPlacesConceptsTextPrinted WrittenImageWords&SoundsVisualsStaticDeepfake Technology CategoriesSource:Voicebot 2023There is a Complexity

23、Spectrum of Deepfake TechnologyAudioSyntheticComplexitySyntheticComplexityVoice ClonesSynthetic VoiceVirtualProximityEfectVideoDigital TwinAugmentationDe-agingVirtual HumanFace SwappingTouch-upVoiceModulationHumanINTRODUCTIONDeepfakes are not represented by a single technology,technique,or modality.

24、Various use cases for replicating human likeness and biometric features involve different levels of technical complexity.In general,the more technical complexity,the lower availability in the market.However,there are exceptions.Inexpensive(or free)and relatively high quality voice clones are availab

25、le to just about anyone today,as are synthetic voices.Another factor to consider is whether the deepfake is applied to recorded media(static or dynamic)or real time communications.The complexity for the latter is higher but it can sometimes be more effective because the other party doesnt have time

26、to contemplate and review the situation.They react to the exchange and may not pick up on the subtle differences between real and fake.They may even think indicators are a reflection of the internet connection or audio channel.PAGE 6 VOICEBOT.AI-All Rights Reserved 2023Deepfake Solution Complexity M

27、apSource:Voicebot 2023INTRODUCTIONPAGE 7 VOICEBOT.AI-All Rights Reserved 2023Voice Clone Awareness is Higher Than Deepfakes,Concern is SimilarExtremelyVeryModerately SlightlyNot At All 7.5%Not Sure 2.3%34.6%Very Concerned57.9%23.3%19.5%12.8%ExtremelyVeryModerately SlightlyNot At All 5.4%Not Sure 2.0

28、%37.6%Very Concerned60.4%22.8%18.8%13.4%DeepfakeVoice Clone020406080100Deepfake and Voice Clone Awareness Level Among ConsumersYesNoNot sure63.6%27.0%YesNoNot sure54.6%34.7%9.4%10.7%DeepfakeVoice CloneDeepfake and Voice Clone Concern Among ConsumersDeepfake and Voice Clone AwarenessAmong ConsumersMo

29、re than half of U.S.adults are aware of deepfakes and voice clones,with the latter registering significantly more recognition.This may not be surprising given that one of the terms is exotic while the other combines easily recognized everyday words.Consumers that are aware of deepfakes and voice clo

30、nes express a great deal of concern with about 60%in both categories saying they were either“very”or“extremely”concerned and over 90%expressing some concern.Source:Voicebot 2023Source:Voicebot 2023The biggest areas of consumer concern relate to areas where sensitive personal identifiable information

31、(PII)is at risk and where false public information could have negative consequences.Notably,“Healthcare”and“Insurance”concerns are significantly outpaced by“Banking.”When it comes to deepfakes concerns,the potential for fraud is top of mind and people associate that with money.Nestled between those

32、staples of PII-heavy consumer services are“Government”and“Media.”Both can be subject to deepfake-laced disinformation and could lead the public to counterproductive actions.Other areas of concern are far lower with only“Home Utilities”breaking the 20%level.And the 6.2%for“Other”suggests consumer ima

33、gination is limited in the scope of threats they may face.PAGE 8 VOICEBOT.AI-All Rights Reserved 2023Domains Where Consumers are Most Concerned About Voice ClonesBanking,Government,and Media Are Top ConcernsINTRODUCTION15.8%Domains Where Consumers are Most Concerned About Voice ClonesNone of the abo

34、veOtherSportsMedia/disinformationInsuranceHealthcareHome UtilitiesGamesPolitics/governmentBanking4.4%6.2%17.1%28.3%48.3%50.1%54.5%54.9%67.5%Source:Voicebot 2023INTRODUCTIONPAGE 9 VOICEBOT.AI-All Rights Reserved 2023Many Consumers Think Companies are Deepfake ReadyWhile consumers express significant

35、concern about deepfakes and voice clones,there is a high degree of uncertainty about the readiness of leading institutions.Around 25%of U.S.consumers say they are“Neutral or Unsure”that businesses are ready for nefarious uses of the technology across key industries.Consumers are most confident that“

36、Banks,Insurance&Healthcare”have already taken steps to protect them against risks.About 40%expressed at least“Somewhat High”confidence.Given the newness of the technology and very recent emergence of detection solutions,it appears consumer expectations may exceed reality.The“Low”confidence categorie

37、s are far higher for“News”and“Social Media.”This is significant as these are channels driving the highest number of encounters.SocialMedia16.3%14.9%14.1%23.2%10.5%8.3%12.7%Extremely LowVery LowSomewhat LowNeutral or UnsureSomewhat HighVery HighExtremely High6.1%9.9%13.9%29.5%16.7%9.2%14.9%Banks,Insu

38、rance&Healthcare11.6%11.4%16.6%26.2%12.6%8.2%13.4%NewsMediaConsumer Confidence in Deepfake and Voice Clone Protection for Key At-Risk IndustriesSource:Voicebot 2023Sentiment Skews Toward the Extremes1.8%22.3%6.2%11.8%10.8%15.6%22.3%1.0%I dont know/not applicableExtremely negativeVery negativeSlightl

39、y negativeNeutralSlightly positiveVery positiveExtremely positive8.8%18.8%6.9%14.9%11.7%15.7%21.6%Voice CloneDeepfake9.9%INTRODUCTIONMost consumer sentiment is driven by the extremes of positive and negative aspects of deepfake and voice clone technologies.In fact,the results look a lot like an inve

40、rse normal distribution.Overall,the unaided positive sentiment for deepfakes was 38.5%to 48.8%negative.The figures were 34.3%positive to 49.1%negative for voice clones.This reflects a measurable negative sentiment bias.Interestingly,consumers that expressed the highest levels of concern were signifi

41、cantly more likely to express positive sentiment about both technologies.Deepfakes registered 51.7%positive and voice clones showed 45.6%among the“Very”and“Extremely”concerned.It appears that more familiarity may lead to both more positive interest and more concern.These consumers better appreciate

42、the benefits and the risks.PAGE 10 VOICEBOT.AI-All Rights Reserved 2023Deepfake and Voice Clone SentimentSource:Voicebot 2023Consumers are most likely to encounter deepfakes and voice clones on social media.The top four responses for both categories were YouTube,TikTok,Instagram,and Facebook.You wil

43、l note the bias toward video on these platforms as YouTube and TikTok encounters were materially higher.Granted all of these platforms have video,but two use the media exclusively.Movies,the news media,and television followed closely behind Facebook and Instagram.Exposure is broad based.However,deep

44、fakes experience exceeds voice clones for all top media sources which suggests that consumers were more likely to experience deepfakes across multiple channels It also may reveal that many people know of voice clones but have not personally encountered them.Consumers were more likely to encounter vo

45、ice clones on audio channels such as Spotify and phone calls.They were also significantly more likely to have created their own voice clone.PAGE 11 VOICEBOT.AI-All Rights Reserved 2023Deepfake Consumer Encounters by ChannelSocial Media Accelerates Deepfake ExposureINTRODUCTIONI have not seen or hear

46、dany deepfakes/voice clonesNot sure/none of the aboveIve created adeepfake/voice cloneOn a callSpotifyVideo gamesSnapchatTelevisionNews mediaMovies/documentariesFacebookInstagramTikTokYouTube21.2%22.6%32.6%33.1%35.4%36.2%37.9%42.8%49.0%7.8%10.9%12.9%14.8%15.0%15.4%21.9%25.7%25.9%26.6%27.5%31.3%39.7%

47、7.3%12.7%11.3%7.6%5.5%3.6%DeepfakeVoice CloneSource:Voicebot 2023INTRODUCTIONPAGE 12 VOICEBOT.AI-All Rights Reserved 2023A Tale of Two OutcomesConsumer sentiment and experience with deepfakes and voice clones serve as a backdrop for these considerations.What are their expectations?What are their fea

48、rs?How will they react to organizational action or non-action to the risks?What do they want?Answers to these questions will influence how enterprises approach the risk posed by these technologies and the level of urgency.The data and conclusions above reflects a small sample of the findings in this

49、 report.On the following pages,you will find additional analysis of consumer sentiment and experience broken down by media consumption,age,income,and gender.You will also find a summary of some high-profile positive and negative examples of deepfake and voice clone use.There is a perfect storm of su

50、pply and demand for deepfakes and voice clones due to a convergence of rising quality and ease of access combined with mass distribution through social and traditional media.The novelty and practical benefits have led to many positive uses along with consumer interest.At the same time,the enhanced r

51、ealism is increasingly embraced by bad actors for nefarious purposes.Deepfakes and voice clones may have introduced the“best of times”for creative pursuits.However,if you are focused on fraud prevention,copyright monitoring,or just want to protect your reputation,are you ready for“the worst of times

52、?”The positive applications will take care of themselves.The real question is how to identify and mitigate the risks posed by negative uses.Deepfake and Voice Clone Scams,Attack Vectors,and Positive UsesSCAMS,VECTORS,AND POSITIVE USESPAGE 14 VOICEBOT.AI-All Rights Reserved 2023The Good,the Bad,and t

53、he Ugly of Deepfakesused the technology to impersonate a corporate officers and steal millions of dollars from a company bank account.They have stolen thousads by impersonating grandchildren.Someone even used a deepfake to impersonate Ukraine President Volodymyr Zelensky and influence the war.The li

54、keness of political figures,celebrities,and family members have all been used to defraud.The problem is likely to get worse as the technologys quality and ease of access rise while costs declines.However,it is not an entirely dire situation.Detection tools are emerging and there are a lot of positiv

55、e use cases proliferating alongside the nefarious.The following pages highlight several recent examples of positive and negative uses of deepfakes and voice clones.You will see that the benefits and risks show up in a variety of forms.Deepfakes and voice clones have shown up in good,bad,and downrigh

56、t ugly circumstances.An important reason that the technology awareness is high is that it has appeared in numerous situations over the past two years.From a positive perspective,major media and social media creators have employed the technology for its entertainment value.Gamers have used it to add

57、more fun to their in-session voice chats and people have employed it to prank friends.Voice clones are also used as prostheses for people that have lost the ability to speak naturally.However,the negative uses are on the rise.Scams distributed on social media and through phone calls are targeting ev

58、eryday consumers.Criminals have SCAMS,VECTORS,AND POSITIVE USESPAGE 15 VOICEBOT.AI-All Rights Reserved 2023Deepfake Simon Cowell Wows AGT FansBackground Metaphysic appeared on Americas Got Talent(AGT)to highlight its deepfake technologys entertainment value.VenueThe solution appeared in three differ

59、ent AGT shows:once with a single singer replicating Cowell,another as a trio including other judges,and a third that brought back an Elvis likeness and voice clone.ResultMetaphysics performance made it to the finals for 2022 and finished the competition as fourth overall.“This is the best act,I beli

60、eve,of the series.”Simon Cowell,AGT CreatorDeepfake technology company Metaphysic entered the Americas Got Talent variety show competition with a replica of show creator and judge Simon Cowell.SCAMS,VECTORS,AND POSITIVE USESPAGE 16 VOICEBOT.AI-All Rights Reserved 2023De-aging Robert De NiroBackgroun

61、d The Irishman chronicled the lives of mafia and union figures over four decades and used a combination of deepfake and CGI technology for de-aging Robert De Niro,Al Pacino,Joe Pesci,and other actors.VenueNetflix movie 2019.ResultThe film received several nominations for best visual effects awards,i

62、ncluding for the Academy Awards.“De-aging is invaluable when you have three actors trying to cover five decades worth of time.”Angela Watercutter,WIREDDeepfake technology was used in The Irishman to digitally de-age Robert De Niro and other actors to portray them at several different ages.SCAMS,VECT

63、ORS,AND POSITIVE USESPAGE 17 VOICEBOT.AI-All Rights Reserved 2023Bringing Andy Warhol to LifeBackground Pop artist Andy Warhol left behind over 20,000 diary pages.These became the source material for The Andy Warhol Diaries documentary and the director wanted to animate the words with a replica of t

64、he artists voice.VenueThe voice clone was used in a six-part Netflix documentary.ResultThe Andy Warhol Diaries received an Emmy nomination for best television documentary in 2022.“When I first heard the AI-created voice,I felt confident it was going to be an incredibly effective way to bring Warhols

65、 diaries to life and humanize the enigmatic artist.”Michael Dayton Hermann,The Andy Warhol FoundationA documentary based on the private diaries of Andy Warhol employed a voice clone to represent the artists thoughts in a 2022 documentary.SCAMS,VECTORS,AND POSITIVE USESPAGE 18 VOICEBOT.AI-All Rights

66、Reserved 2023Voice Clone Fraud Used to Steal$35 MillionThe Scam A Hong Kong branch manager received a call from someone impersonating the Japanese companys director using a voice clone.He was asked to authorize a bank transfer for an acquisition.The manager believed it was legitimate because he reco

67、gnized the voice and it matched a request in hacked email correspondence.Distribution VectorVoice clone used in a phone call and confirmed by emails.ImpactStolen funds,totaling$35 million,were traced to at least 17 individuals,and sent to bank accounts worldwide.“Audio and visual deep fakes.pose a h

68、uge threat to data,money and businesses.”Jake Moore ESETA voice clone of a company director was employed to convince a branch manager to authorize the transfer of$35 million for an“acquisition”that was actually a scam.SCAMS,VECTORS,AND POSITIVE USESPAGE 19 VOICEBOT.AI-All Rights Reserved 2023Deepfak

69、e Fraud Imitating a Consumer Advocate The Scam A video featuring a deepfake of Martin Lewis,the well known UK consumer advocate,discussing an investment in Quantum AI,falsely linked to Elon Musk.The videos realism made it convincing to viewers.Distribution VectorThe video initially appeared on Faceb

70、ook and was reported on Instagram and Twitter.ImpactLewis criticized the lack of social media platforms safeguards that allow deepfake scams to thrive.Total losses from the scam are unkown.“This is a deepfakewere not quite sure of the exact tech.We are in a dangerous dystopian future,and nothing is

71、being done to protect people.”A deepfake of Martin Lewis,a trusted consumer advocate in the UK,was used in a video shared on social media to promote a fake app and investment opportunity supposedly associated with Elon Musk.SCAMS,VECTORS,AND POSITIVE USESPAGE 20 VOICEBOT.AI-All Rights Reserved 2023D

72、eepfake Fraud Imitating a Popular YouTube Creator The Scam A deepfake ad featuring a digital twin of MrBeast asked viewers to click a link and sign up for a$2 iPhone 15 Pro giveaway.Distribution VectorThe deepfake ad was initially identified on social media platforms,including Facebook,Instagram,and

73、 Twitter.ImpactIt is unclear how many people fell for the scam.The incident prompted discussions about deepfake risks.“This highlights the urgent need for robust measures against deceptive deepfake content.Were navigating uncharted territory,and platforms must prioritize protective mechanisms to cou

74、nter the risks.”MrBeastA deepfake of popular Youtuber MrBeast(Jimmy Donaldson)was used in a video shared on social media inducing followers to sign up for a fraudulent iPhone giveaway.SCAMS,VECTORS,AND POSITIVE USESPAGE 21 VOICEBOT.AI-All Rights Reserved 2023Voice Clone Fraud Imitating a Grandchild

75、The Scam A call is made by an impersonator using a voice clone of a grandchild that claims to be in trouble and in need of money quickly for bail,to pay a lawyer,or to pay-off kidnappers.Cash in then typically exchanged.Distribution VectorPhone call.ImpactA recent series of incidents in Canada led t

76、o more than$40,000 in losses by duped grandparents in one area and over$200,000 in another.“Youd do anything for your grandchildren.And they took advantage of that.”Grandmother duped in scam.”as reported by CBCVoice clones are used to imitate a loved one or friend on a phone call and ask for money t

77、o get them out of jail,pay for a tow truck,or save them from a kidnapper.SCAMS,VECTORS,AND POSITIVE USESPAGE 22 VOICEBOT.AI-All Rights Reserved 2023Deepfake Disinformation Attempts to Influence a WarThe Disinformation A relatively low-quality deepfake of Ukraine president Volodymyr Zelensky sought t

78、o sow disinformation by making the military believe that the country had surrendered to Russia.Distribution VectorHackers published it on a Ukrainian news website and social media.It also appeared in a Ukrainian TV news broadcast.ImpactThe deepfake was quickly identified and removed.“It pollutes the

79、 information ecosystem,and it casts a shadow on all content,which is already dealing with the complex fog of war.”Hany Farid,UC BerkeleyA deepfake of Volodymyr Zelensky portrayed the Ukraine President as instructing the military and citizens to lay down their arms in the conflict with Russia.Deepfak

80、e ConsumerAwareness&Sentiment Over half of U.S.adults report awareness of deepfake technology,with just over one-third saying they have not heard of it and another 10.6%unsure.This awareness is not necessarily grounded in a common definition or common sentiment,as you will see in this report.However

81、,it does suggest that what was just recently an obscure AI technology has quietly gained at least surface-level awareness among consumers.The emergent nature of consumer exposure to deepfakes,combined with socio-economic and gender differences,has also led to divergent views on the merits of the tec

82、hnology.PAGE 24 VOICEBOT.AI-All Rights Reserved 2023Deepfake is No Longer an Obscure Technical TermDEEPFAKE CONSUMER AWARENESS&SENTIMENTDeepfake Awareness Among ConsumersDeepfake Awareness Among Consumers10.6%Im notsure34.7%No54.7%YesSource:Voicebot 2023 Men report awareness of deepfake technology a

83、t far higher rate than women.Over 66%of men said they were familiar with deepfakes compared to only 42%of women.Men are more concerned than women about deepfakes.They are also 50%more likely to have created a deepfake than women,though both figures are less than 10%of the group that knows of deepfak

84、es and less than 4%of the total U.S.adult population.Beyond awareness and deepfake creation,men were more likely to have encountered a deepfake on YouTube(52.4%to 46.7%),while women had the edge over men in Tiktok(43.1%to 38.4%).The other media channels were within the margin of error for the group

85、differences.PAGE 25 VOICEBOT.AI-All Rights Reserved 2023Deepfake Awareness by GenderMen Report Higher Deepfake AwarenessDEEPFAKE CONSUMER AWARENESS&SENTIMENTDeepfake Awareness by GenderMen66.1%Women42.2%All 54.7%All 54.7%Source:Voicebot 2023Deepfake Awareness Among Select Media ConsumersStar Wars Mo

86、vies/ShowsAmericasGot TalentAndy WarholDiariesAllIndiana Jones Dial of DestinyNegativeNeutral+Dont KnowPositive75.9%68.3%63.8%56.1%54.7%Deepfake awareness is highest among consumers that have viewed popular media where the technology was featured.Three out of four viewers of the“Andy Warhol Diaries”

87、documentary report deepfake awareness compared to just 56%of“Americas Got Talent”viewers and 54.7%overall.The voice clone storyline was inescapable in the Warhol production,while Americas Got Talent(AGT)only featured a deepfake in the 2022 season.Demographic factors may also explain this difference,

88、given the profile of documentary viewers.“Star Wars”and“Indiana Jones”have broader appeal than documentaries and also created unavoidable recognition of deepfakes through de-aging and recreating deceased actors.PAGE 26 VOICEBOT.AI-All Rights Reserved 2023Deepfakes in Entertainment Drive Awareness DE

89、EPFAKE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023 The top four channels for deepfake exposure are social media platforms:YouTube,TikTok,Instagram,and Facebook,with 49.0%,42.8%,37.9%,and 36.2%reported encounters,respectively.These are followed by three mass media channels:movies/documentaries(3

90、5.4%),news media(33.1%),and television(32.6%).Despite some high-profile deepfake songs,Spotify was only identified by 12.7%of consumers.Surprisingly,over 11%of consumers said they had experienced a deepfake on a call,and nearly 8%have some experience personally creating a deepfake.PAGE 27 VOICEBOT.A

91、I-All Rights Reserved 2023Deepfake Consumer Encounters by ChannelSocial Media Drives the Most Deepfake ExposureDEEPFAKE CONSUMER AWARENESS&SENTIMENTI have not seen or heard any deepfakesNot sure/none of the aboveIve created a deepfakeOn a callSpotifyVideo gamesSnapchatTelevisionNews mediaMovies/docu

92、mentariesFacebookInstagramTikTokYouTube5.5%7.6%11.3%12.7%21.2%22.6%32.6%33.1%35.4%36.2%37.9%42.8%49.0%3.6%Source:Voicebot 2023Consumer Definitions of Deepfake TechnologyUsing AI to replace the face of oneperson with another in a videoUsing AI to impersonate the voiceof someone elseFace-swappingVoice

93、 cloneA tool used by criminals tocommit bank fraudA technology used to makeactors look younger in filmsA technology used to makefunny videosNone of the above68.4%66.7%52.7%51.1%42.7%42.2%32.9%0.8%Consumer Definitions of Deepfake TechnologyNegativeNeutral+Dont KnowPositivePAGE 28 VOICEBOT.AI-All Righ

94、ts Reserved 2023How Consumers Define DeepfakesDEEPFAKE CONSUMER AWARENESS&SENTIMENT While“deepfake”is a term often associated with video,and its origin was strictly related to video,consumers also equate it with voice cloning.About 68.4%think of deepfake as replacing a humans face with another,and 6

95、6.7%believe it involves using AI to impersonate someones voice.Over 50%apply the term to both media modes.We see similar trends for video and voice with the terms“face-swapping”and“voice clone,”which were reported by 52.7%and 51.1%respectively.Nearly 43%indicated that deepfakes are used to commit ba

96、nk fraud,while a similar percentage knew they are employed to de-age actors in films.Source:Voicebot 2023Deepfake Sentiment Among Consumers22.3%9.9%6.2%11.8%10.8%15.6%22.3%1.0%I dont know/not applicableExtremely negativeVery negativeSlightly negativeNeutralSlightly positiveVery positiveExtremely pos

97、itive38.4%Positive48.7%NegativePAGE 29 VOICEBOT.AI-All Rights Reserved 2023Deepfake Sentiment Biases Toward the ExtremesDEEPFAKE CONSUMER AWARENESS&SENTIMENT Consumer sentiment towards deepfake skews negative,but the extremes for positive and negative associations are identical:22.3%of consumers are

98、 extremely positive about deepfakes while another 22.3%are extremely negative,indicating that deepfakes are a polarizing topic.Overall,positive sentiment lags behind negative sentiment by over 10%,with 38.4%having at least slightly positive feelings about the technology compared to 48.7%that are neg

99、ative.The gap narrows slightly to 5.7%when focusing solely on“very”and“extremely”responses.With only 11.8%claiming neutrality and polarization at the extremes,deepfakes are poised to become a contentious topic.Source:Voicebot 2023 The polarization of deepfake sentiment varies across consumers that h

100、ave watched different types of media.Documentary viewers of The Andy Warhol Diaries show a deepfake positive sentiment totalling 71.1%,followed by the most recent Indiana Jones movies at 56.7%.This suggests that viewers of those media saw tangible value in the technology.By contrast,AGT and Star War

101、s viewers were only slightly more positive than the general population.It appears that exposure to deepfakes in media has an effect on sentiment,but it can range from negligible to extreme.PAGE 30 VOICEBOT.AI-All Rights Reserved 2023Deepfake Sentiment Among Media ConsumersDeepfakes in Entertainment

102、Raises Positive SentimentDEEPFAKE CONSUMER AWARENESS&SENTIMENTNegativeNeutral+Dont KnowPositivePositiveNegativeNeutralAndy Warhol DiariesIndiana Jones Dial of DestinyStar Wars Movies/ShowsAmericas Got TalentAll12.8%48.8%38.5%13.0%46.0%41.0%12.9%45.3%*+31.2%*-3.4%*Positivity Score*+53.6%*-5.0%*-10.0%

103、41.8%56.7%71.1%17.7%25.5%11.4%17.5%Source:Voicebot 2023PAGE 31 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE CONSUMER AWARENESS&SENTIMENTWhat Consumers Like About DeepfakesWhat Consumers Like About Deepfakes22.3%Not sureThe creativityThey improve realism in entertainmentThey are funny or entertaining

104、Nothing39.4%10.1%29.7%27.7%40.0%What Consumers Like About Deepfakes Just over one-quarter of consumers find that deepfakes have no redeeming qualities,while over 60%were able to identify features they liked about the technology.Consumers most often cited positive deepfake traits in terms of what the

105、y enable beyond existing technologies or their pure entertainment value.Only 29.7%of U.S.adults appreciated deepfakes for improving realism in entertainment.Interestingly,despite being a technology based on fakery,there are alternative methods for creating realism.The citation for“creativity”at 39.4

106、%suggests that the unique attributes of deepfakes and the new options they provide are stronger drivers of interest.Source:Voicebot 2023PAGE 32 VOICEBOT.AI-All Rights Reserved 2023Deepfake Concern Level Among ConsumersDeepfake Concern is High Among ConsumersDEEPFAKE CONSUMER AWARENESS&SENTIMENTExtre

107、melyVeryModerately SlightlyNot At All 5.4%Not Sure 2.0%37.6%Very Concerned60.4%22.8%18.8%13.4%Consumers often express concern about online privacy while using social media and other apps that seemingly undermine those concerns.This paradox also seems to apply to deepfakes.Consumers can like a techno

108、logy while still being concerned about its risks.Over 90%of consumers have at least some concern about deepfakes,with 80%being at least moderately concerned.A total of 60.4%are extremely or very concerned.Yet,nearly 40%express positive sentiment toward the technology.This suggests that many consumer

109、s see benefits and likely presume the risks can be managed.Source:Voicebot 2023Deepfake Sentiment Among Consumers with High Concern About Personal RisksNegativePositiveNegativeNeutral+Dont KnowPositive38.5%51.7%48.8%20.6%ALLExtremely+Very ConcernedPAGE 33 VOICEBOT.AI-All Rights Reserved 2023The Most

110、 Concerned are Also the Most PositiveDEEPFAKE CONSUMER AWARENESS&SENTIMENT Surface-level consumer dissonance between concern and interest also shows up in the sentiment data.Consumers who said they were extremely or very concerned about deepfake technology impacting them personally also had positive

111、 sentiment at a far higher rate than average.Similarly,the negative sentiment was much lower.How do we explain this?It appears that greater interest in the technology also leads to higher perceived benefits and concerns.The most concerned consumers also cited benefits such as improved realism in mov

112、ies and entertainment value of the technology at a higher rate than the general population.Source:Voicebot 2023Consumers That Know Someone Who Has Had a Deepfake Made of ThemConsumers That Know Someone Who Has Had a Deepfake Made of Them49.3%I dontknow31.3%No19.4%YesU.S.Adults7.1%I dont know57.4%No3

113、5.6%YesAware ofDeepfakesPAGE 34 VOICEBOT.AI-All Rights Reserved 2023Personal Connection to Deepfakes are SignificantDEEPFAKE CONSUMER AWARENESS&SENTIMENT Over one-third of U.S.adults familiar with deepfakes report knowing someone who has had a deepfake made about them.As expected,this figure exceeds

114、 that for voice clones(31%),as“deepfake”is a broader category.Importantly,those reporting deepfakes of someone they know may be referring to voice clones,video manipulation,or other techniques.From a total population standpoint,just under 20%say they know someone with a deepfake.Approximately half o

115、f the population is unsure if an acquaintance has had a deepfake created about them.Source:Voicebot 2023 While concern about using deepfakes for fraud or other personal harm is high,four out of five people who know someone with a deepfake made said it was done with their consent.By contrast,13.2%sai

116、d the deepfake was made without their acquaintances consent,and another 5.9%know people with multiple deepfakes,some with and some without consent.That means about 19%of U.S.adults familiar with deepfakes know someone that has had a deepfake made without their consent.When considered against the tot

117、al population the non-consent figure is 3.7%,representing over 12 million U.S.adults that know someone that faced this problem.PAGE 35 VOICEBOT.AI-All Rights Reserved 2023Percent of Deepfakes Consumers Say Were Made with Persons ConsentDeepfakes Made without Consent is SignificantDEEPFAKE CONSUMER A

118、WARENESS&SENTIMENTPercent of Deepfakes Consumers Say Were Made with Persons Consent13.2%No80.0%Yes,it was made with their consent5.9%Some made with consent and some without0.9%I dont knowSource:Voicebot 2023PAGE 36 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE CONSUMER AWARENESS&SENTIMENTDeepfake Con

119、cern Level for Consumers That Know Someone That Has a Deepfake18.8%22.8%13.4%5.4%2.0%37.6%9.2%17.7%68.6%0.5%1.4%2.6%+31.0%-5.1%-9.6%-10.8%-4.0%-1.6%Not sureNot at allSlightlyModerately HighVery HighExtremely HighAllHad Deepfake Made About Them or Someone They KnowDiferenceWhat Consumers Like About D

120、eepfakes A personal connection to someone with a deepfake correlates with a higher level of concern about the technology.While 60.4%of all U.S.adults aware of deepfakes report extreme or high concern about negative impact,that figure rises to 86.3%if they know someone with a deepfake made about them

121、.However,a personal connection to someone with a deepfake results in more positive sentiment about deepfakes compared to the general population.This is true whether the deepfake was made with or without consent.Greater awareness seems to translate into more positive sentiment.Source:Voicebot 2023Voi

122、ce Clone ConsumerAwareness&SentimentVoice Clone Awareness Among Consumers020AndroidVoice Clone Awareness Among Consumers27.0%No63.6%Yes2.8%Im not sure Nearly two-thirds of U.S.adults say they are aware of the term“voice clones,”which is nearly percentage points higher than for the term“deepfake.”A s

123、lightly higher number claimed awareness but could not provide accurate definitions.However,consumers disagree on what constitutes a voice clone,and some people choosing incorrect definitions were excluded from the data in related responses.Nevertheless,the term is widely recognized due to its presen

124、ce in social media,traditional media,games,and even in phone calls.PAGE 38 VOICEBOT.AI-All Rights Reserved 2023Most Consumers Are Aware of Voice ClonesVOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023Voice Clone Awareness by GenderMen71.5%Women60.0%All 65.8%All 65.8%Men Report Higher Voic

125、e Clone AwarenessVoice Clone Awareness by Gender Similar to the term“deepfake,”men report higher awareness of voice clone technology than women.Over 71%of men reported familiarity with voice clones,compared to 60%of women.Men are also more concerned than women about deepfakes and are 50%more likely

126、to have created one,though both figures are under 10%among those aware of deepfakes and under 4%of the total U.S.adult population.Beyond awareness and deepfake creation,men were more likely to have encountered a deepfake on YouTube(52.4%to 46.7%),while women had the edge over men in Tiktok(43.1%to 3

127、8.4%).The other media channels were within the margin of error for the group differences.PAGE 39 VOICEBOT.AI-All Rights Reserved 2023VOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023PAGE 40 VOICEBOT.AI-All Rights Reserved 2023Voice Clone Awareness Higher Among Select MediaVOICE CLONE CONS

128、UMER AWARENESS&SENTIMENTVoice Clone Awareness Among Select Media ConsumersVoice Clone Awareness Among Select Media ConsumersStar Wars Movies/ShowsAmericasGot TalentAndy WarholDiariesAllIndiana JonesDial of DestinyNegativeNeutral+Dont KnowPositive79.3%75.0%74.3%70.0%65.8%U.S.adults who have viewed me

129、dia prominently featuring voice clones show higher awareness of the technology.Americas Got Talent(AGT)featured an act over the summer of 2022 that included an Elvis voice clone in the finals,boosting awareness by about 4%.Star Wars used tools to recreate the voices of deceased actors and de-age oth

130、ers,contributing to an 8%higher awareness.The Andy Warhol Diaries made voice clones integral to its production,with a Warhol voice clone narrating from his official diaries.Viewers of this documentary recognized the term“voice clone”at a rate of more than 13 percentage points higher than average.Sou

131、rce:Voicebot 2023Voice Clones are Encountered Across Media FormatsPAGE 41 VOICEBOT.AI-All Rights Reserved 2023VOICE CLONE CONSUMER AWARENESS&SENTIMENT The top two channels for voice clone exposure are the same as for deepfakes:YouTube and TikTok at 39.7%and 31.3%,respectively.These are followed by t

132、he three traditional mass media channels:Television(27.5%),Documentaries and Movies(26.6%),and News Media(25.9%),which edged out the social media heavyweights Facebook,Instagram,and Snapchat.Despite some high-profile deepfake songs,Spotify was identified by only 12.7%of consumers for voice clone enc

133、ounters.Surprisingly,over 15%of consumers have experienced a voice clone on a call,and over 7%have some experience personally creating one.Deepfake Consumer Encounters by ChannelI have not heard any voice clonesNot sure/none of the aboveIve created a voice cloneOn a callSpotifyMuylti player video ga

134、mesSnapchatTelevisionNews mediaMovies/documentariesFacebookInstagramTikTokYouTube7.8%10.9%12.9%14.8%15.0%15.4%21.9%25.7%25.9%26.6%27.5%31.3%39.7%7.3%Source:Voicebot 2023Consumer Definitions of Voice Clone Technology19.4%Consumer Definitions of Voice Clone TechnologyNot sure/none of the aboveA fun fe

135、ature for gamingAnother term for autotunetechnologyused by singersUsed by criminals to commitbank fraudIt is a voice changerA recording of anindividuals voiceUsed by people that cannot speaknaturally due to a health problemUsed by comedians toimpersonate a celebrityUsing AI to recreate the voice ofs

136、omeone that is deceasedtUsing AI to impersonate thevoice of someone else2.2%18.5%24.1%26.8%30.2%32.2%45.0%50.7%68.9%Most consumers aware of the term“voice clone”associate it with AI used to impersonate a voice(68.9%)or recreate the voice of someone that is deceased(50.7%).The third most reported def

137、inition,at 45.0%was that voice clones are“used by criminals to commit bank fraud.”Consumers that chose this response were also more likely to be extremely or very concerned about voice clones.Two incorrect responses,“a recording of an individuals voice”and“another term for autotune,”led to the exclu

138、sion of respondents who chose only these definitions.In addition,over 30%of people who chose a correct definition also selected an incorrect option,suggesting there is an inconsistent understanding of what constitutes a voice clone.PAGE 42 VOICEBOT.AI-All Rights Reserved 2023How Consumers Define Voi

139、ce ClonesVOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 202334.3%Positive49.0%NegativeVoice Clone Sentiment Among Consumers8.8%18.8%6.9%14.9%11.7%15.7%21.6%1.8%I dont know/not applicableExtremely negativeVery negativeSlightly negativeNeutralSlightly positiveVery positiveExtremely positiveVo

140、ice Clone Sentiment Biases Toward the ExtremesVoice Clone Sentiment Among Consumers Similar to the data on deepfakes,consumer sentiment about voice clones skews toward extremes.The two most selected responses were“extremely negative”(21.6%)and“extremely positive”(18.8%).Overall,positive sentiment to

141、taled just over one-third(34.3%),while negative sentiment was just under one-half(49.0%).More than half of the positive sentiment(55%)falls under the“extremely”category,whereas negative sentiment is more distributed,with“extremely”accounting for only 44%of the negative responses.PAGE 43 VOICEBOT.AI-

142、All Rights Reserved 2023VOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023Voice Clone Sentiment Among Media ConsumersVoice Clone Sentiment Among Media ConsumersNegativeNeutral+Dont KnowPositivePositiveNegativeNeutralAndy Warhol DiariesIndiana Jones Dial of DestinyStar Wars Movies/ShowsAmer

143、icas Got TalentAll16.6%49.1%34.3%17.7%46.9%35.4%16.1%43.2%*+22.5%*+1.6%*Positivity Score*+46.7%*-6.9%*-9.9%40.7%51.2%69.0%17.7%31.2%12.4%18.5%Positive voice clone sentiment is correlated with exposure to media that showcases the technology,though it is not clear how big of a factor it is.AGT include

144、d a voice clone in the finals in 2022,but not every audience member saw those episodes and the positive sentiment was similar to the overall group.By contrast,viewers of the Andy Warhol Diaries show a strongly positive sentiment totalling 69.0%,followed by the most recent Indiana Jones movies at 51.

145、2%.The former used a voice clone;the latter did not.Star Wars used voice clones for movies and the Boba Fett series,and its viewers were significantly more positive about the technology than the general population.Similar to deepfakes,it appears exposure to voice clones in media may affect sentiment

146、,but it can range from negligible to extreme.PAGE 44 VOICEBOT.AI-All Rights Reserved 2023Voice Clone Exposure in Media Raises Positive SentimentVOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 202335.6%ThecreativityUsing them in multiplayer video gamesFoolingfriendsThey improve realism in ent

147、ertainmentThey are funny or entertaining22.2%Nothing30.9%Not sure8.7%23.2%30.5%26.0%Consumers Like Voice Clones for Comedy and GamingWhat Consumers Like About Voice Clones Consumers express strong interest in voice clones.The most prominent postive reaction related to using of voice clones in multip

148、layer video games.Similar to deepfakes,a significant number of U.S.adults do appreciate the creativity and enhanced realism provided by voice clones,but at a far lower rate:39.4%to 22.2%and 29.7%to 23.2%,respectively.About one-quarter of survey participants that know about voice clones said they lik

149、e using them to fool friends,while 30.9%see no redeeming qualities of the technology.PAGE 45 VOICEBOT.AI-All Rights Reserved 2023VOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023Voice Clone Concern is High Among ConsumersPAGE 46 VOICEBOT.AI-All Rights Reserved 2023VOICE CLONE CONSUMER AWA

150、RENESS&SENTIMENTVoice Clone Concern Level Among ConsumersVoice Clone Concern Level Among ConsumersExtremelyVeryModerately SlightlyNot At All 7.5%Not Sure 2.3%34.6%23.3%19.5%12.8%Very Concerned59.7%Over 90%of U.S.adults express some concern that voice clones could negatively impact them personally,wi

151、th about 80%being at least moderately concerned,and 60%indicating extreme or very high concern.Voice clones may have many meritorious uses,but that has not overridden latent concerns.This concern is also expressed at a time when fraudulent use of voice clones is on the rise,particularly to fool fami

152、ly members about a child or relative that needs help.Source:Voicebot 2023Voice Clone Sentiment Among Consumers With High Concern About Personal RisksVoice Clone Sentiment Among Consumers with High Concern About Personal RisksNegativePositiveNegativeNeutral+Dont KnowPositive34.3%45.6%49.1%49.2%Extrem

153、ely+Very ConcernedALL Consumers that are most concerned about the personal risks posed by voice clones expressed about the same strong negative sentiment as the total survey sample.However,these“most concerned”were also more positive than the overall average.This was not as extreme of an inversion a

154、s with deepfakes,but it is notable.As with deepfakes,it appears that greater interest in voice clone technology may lead to higher perceived value even if the concern level remains significant.The most concerned consumers were also significantly more likely to cite the benefits of voice clones in mu

155、ltiplayer video games and for greater realism in entertainment than the overall average.Of course,they were also more likely to know about fraudulent criminal use.PAGE 47 VOICEBOT.AI-All Rights Reserved 2023The Most Concerned are More PositiveVOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2

156、02339%NoI dont know40%21%Yes9.1%I Dont KnowU.S.Adults59.4%No31.5%YesHad a Voice Clone Made of ThemKnow ofVoiceClonesPersonal Knowledge of Voice Clones is WidespreadConsumers That Know Someone Who Has Had a Voice Clone Made of Them From a total population standpoint,about 21%of U.S.adults say they kn

157、ow someone who has had a voice clone made.Approximately 40%of the population either does not know if an acquaintance has had a voice clone made or is unaware of the term,and 39%answered“no”.About one-third of U.S.adults familiar with voice clones say they know someone who has had a clone made.This f

158、igure is slightly lower than for deepfakes(35.6%),which is a broader category.PAGE 48 VOICEBOT.AI-All Rights Reserved 2023VOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023Percent of Voice Clones Consumers Say WereMade With Persons ConsentPercent of Voice Clones Consumers Say Were Made wit

159、h Persons Consent11.1%No82.9%Yes,it was made with their consent4.4%Some made with consent and some without1.6%I dont know Over 80%of U.S.adults that know someone who had their voice cloned using AI said the process was done with consent.The concern about voice clone use for fraud reflecting personal

160、 risk is likely influenced by the 15.5%that said they know of voice clones made without consent.While those with knowledge of acquaintances having voice clones made without consent represent only 3.2%of U.S.adults,that reflects a total of nearly 11 million people.Voice cloning technology is becoming

161、 easier to access,less expensive,and more widely used by consumers as well as enterprises.These figures,both for consent and non-consent,are sure to rise.PAGE 49 VOICEBOT.AI-All Rights Reserved 2023Most Voice Clones Are Made with ConsentVOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 202319.

162、5%23.3%12.8%7.5%2.3%34.6%Not sureNot at allSlightlyModerately HighVery HighExtremely HighAll10.6%19.5%65.0%0.2%1.3%3.3%Had Voice Clone Made About Them or Someone They KnowVoice Clone Concern Level for Consumers That Know Someone That Has a Voice CloneDifference+30.4%-3.3%-8.9%-9.5%-6.2%-2.1%Voice Cl

163、one Concern is Higher When You Know a VictimVoice Clone Concern Level for Consumers That Know Someone That Has a Voice Clone Similar to the data for deepfakes,a personal connection with someone that has had a voice clone made is correlated with a higher level of concern about the technology.While 57

164、.9%of U.S.adults that are aware of voice clones report“extremely”or“very”high concern about personal negative impacts,the figure rises to 84.5%if they know someone with a voice clone.Most of that rise is attributed to“extremely high”which is more than 30 points higher than the average.Interestingly,

165、a personal connection to someone with a voice clone results in higher positive sentiment about the technology.This is true whether or not the voice clone was made with their consent.Higher awareness is correlated with more positive sentiment and concern.PAGE 50 VOICEBOT.AI-All Rights Reserved 2023VO

166、ICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023Consumer Interest in Voice Authentication Interest in Voice Authentication Among Those Very Concerned About Voice ClonesConsumer Interest in Voice Authentication 15.0%Not at allinterested2.8%Im not sure11.5%Not veryinterested19.3%Neutral27.3%

167、Veryinterested24.0%Somewhatinterested80100120Very Concerned About Voice Clones12.3%Not at allinterested1.2%Im not sure11.5%Not veryinterested9.3%Neutral52.0%Veryinterested16.8%Somewhatinterested While nearly 60%of U.S.adults rate themselves as very or extremely concerned about voice clones and over

168、90%have at least some concern,over half are interested in using voice authentication when interacting with a bank,healthcare provider,or insurance company.Among consumers that are most concerned about voice clone risk,the interest in voice authentication climbs to 69%.Consumers are concerned voice c

169、lones may be used to harm them,but appear to have confidence that voice authentication can combat the risk.PAGE 51 VOICEBOT.AI-All Rights Reserved 2023Consumers Show Strong Interest in Voice AuthenticationVOICE CLONE CONSUMER AWARENESS&SENTIMENTSource:Voicebot 2023Source:Voicebot 2023Comparing Deepf

170、ake and Voice Clone Consumer Experience and Sentiment The term voice clone registered higher consumer awareness than deepfake,at 63.6%to 54.6%,respectively.This may be surprising when you consider how much higher deepfake registers in Google trends data.However,an explanation may be that the meaning

171、 of voice clone is clear because most people are well aware of cloning.By contrast,deepfake is a portmanteau that originated in 2017,and“deep”stands for deep learning,a term few U.S.adults have encountered outside of AI discussions.As a result,deepfake could drive more searches to learn the meaning.

172、Another consideration is that 83%of U.S.adults that know about deepfakes also know about voice clones versus just 69%of U.S.adults with voice clone awareness know of deepfakes.PAGE 53 VOICEBOT.AI-All Rights Reserved 2023Voice Clones Have Higher Awareness Than DeepfakesCOMPARING DEEPFAKE AND VOICE CL

173、ONEDeepfake and Voice Clone Awareness Among ConsumersYesNoNot sure63.6%27.0%YesNoNot sure54.6%34.7%9.4%10.7%DeepfakeVoice CloneSource:Voicebot 202355.0%31.2%61.8%65.2%All 54.6%All 63.8%68.2%67.1%63.1%52.8%61+45-6030-4418-29DeepfakeVoice Clone Deepfake and voice clone awareness declines gradually as

174、age cohorts rise up to 60 years,and then falls off precipitously.The decline is more extreme for deepfakes.While the difference between the 18-29 and 45-60 cohorts is just over four percentage points for voice clones,it is nearly 10 percentage points for deepfakes.Similarly,deepfake awareness drops

175、by twenty-four percentage points between the 45-60 and the 61+age cohorts,while it is only about ten percentage points for voice clones.Voice clone awareness is higher in each cohort,and the gap widens as age increases.PAGE 54 VOICEBOT.AI-All Rights Reserved 2023Deepfake and Voice Clone Awareness by

176、 AgeAwareness Declines as Age IncreasesCOMPARING DEEPFAKE AND VOICE CLONE Source:Voicebot 2023Deepfake and Voice Clone Awareness by Income Deepfake and voice clone awareness rises with income.While three-quarters of U.S.adults with income over$125,000 know about voice clones and more than two-thirds

177、 know of deepfakes,those figures are just 56.5%and 43.6%,respectively,for income less than$50,000.The awareness gap between the terms is also greatest in the lowest income cohort.As income rises,the likelihood of a consumer knowing either or both terms rises.PAGE 55 VOICEBOT.AI-All Rights Reserved 2

178、023Awareness Rises with IncomeCOMPARING DEEPFAKE AND VOICE CLONEDeepfake and Voice Clone Awareness by Income$125k+$50k-$125k43.6%57.1%68.0%All 54.6%All 65.8%75.0%63.9%56.6.1%$0-$50kDeepfakeVoice CloneSource:Voicebot 2023 U.S.adults report they are“extremely”or“very”concerned about negative impacts a

179、t a higher rate for deepfakes than voice clones.The total for deepfakes is 60.4%and 57.9%for voice clones.The key difference shows up in the“extremely”category where deepfake outpaced voice clones by three percentage points.All of the other concern categories vary by less than one percentage point e

180、xcept for“not at all.”7.5%of consumers that are familiar with voice clones express no concern compared to 5.4%of those with knowledge of deepfakes.Granted,these figures are both very small.Over 90%of consumers that know of these technologies harbor at least some concern.PAGE 56 VOICEBOT.AI-All Right

181、s Reserved 2023Deepfake and Voice Clone ConcernAmong ConsumersConsumers are Slightly More Concerned About DeepfakesCOMPARING DEEPFAKE AND VOICE CLONEExtremelyVeryModerately SlightlyNot At All 7.5%Not Sure 2.3%34.6%Very Concerned57.9%23.3%19.5%12.8%ExtremelyVeryModerately SlightlyNot At All 5.4%Not S

182、ure 2.0%37.6%Very Concerned60.4%22.8%18.8%13.4%DeepfakeVoice CloneSource:Voicebot 2023Deepfake&Voice Clone Concern Level Among Consumers by Income46.7%76.5%61.5%All 60.4%All 57.9%45.8%70.1%61.8%$125k+$50k-$125k$0-$50kDeepfakeVoice ClonePAGE 57 VOICEBOT.AI-All Rights Reserved 2023The Highest Levels o

183、f Concern Rise with IncomeCOMPARING DEEPFAKE AND VOICE CLONE Consumers with the highest incomes also express the highest level of concern.Less than half of consumers with income below$50,000 said they were“extremely”or“very”concerned about the potential of deepfakes or voice clones to negatively imp

184、act them personally.Those figures rise to over 70%for U.S.adults with annual income over$125,000.Interestingly,you will see in the next section that higher income is also associated with higher expectations that companies are taking steps to protect their customers against deepfakes and voice clones

185、.Source:Voicebot 2023Deepfake Sentiment Among Consumers61.9%60.4%54.5%60.8%57.9%MenAllWomen57.1%Extremely+Very Concerned Voice CloneExtremely+Very Concerned DeepfakePAGE 58 VOICEBOT.AI-All Rights Reserved 2023Men Report More Concern Than WomenCOMPARING DEEPFAKE AND VOICE CLONE Men report significant

186、ly more concern than women about the potential of deepfakes and voice clones to cause negative personal impact.61.9%of men had“extremely”or“very”high concern about deepfakes compared to 57.1%of women.The gap was wider for voice clones with 60.8%for men and 54.5%for women.These results may be surpris

187、ing to anyone familiar with the rise of deepfake revenge porn and its disproportionate impact on women.However,it appears that concern has not elevated the perceived risk for women over men.In fact,around 7%of women expressed no concern at all for personal negative impact from either technology comp

188、ared to about 5%for men.Source:Voicebot 2023 Deepfakes are encountered more broadly than voice clones on social media,in traditional mediasuch as television,movies,and the news and in video games.This is not surprising as voice clones are a subset of the broader deepfake category,and these are gener

189、ally visual-first channels.In addition,it corresponds with the finding that consumers say they are about twice as likely to have no prior exposure to a voice clone(7.3%)as a deepfake(3.6%).Voice clones outpaced deepfake exposure in the audio-first channels,such as Spotify and a phone call.Exposure i

190、s also higher for creating a voice clone(10.9%)compared to a deepfake(7.6%).The rise of free voice cloning sites and ease of creation is the likely cause of this difference.PAGE 59 VOICEBOT.AI-All Rights Reserved 2023Deepfake and Voice Clone Consumer Encounters by ChannelDeepfakes are Experienced Ac

191、ross More ChannelsCOMPARING DEEPFAKE AND VOICE CLONEEncounters by ChannelI have not seen or heardany deepfakes/voice clonesNot sure/none of the aboveIve created adeepfake/voice cloneOn a callSpotifyVideo gamesSnapchatTelevisionNews mediaMovies/documentariesFacebookInstagramTikTokYouTube21.2%22.6%32.

192、6%33.1%35.4%36.2%37.9%42.8%49.0%7.8%10.9%12.9%14.8%15.0%15.4%21.9%25.7%25.9%26.6%27.5%31.3%39.7%7.3%12.7%11.3%7.6%5.5%3.6%DeepfakeVoice CloneSource:Voicebot 2023Deepfake and Voice CloneConcern by DomainDomains Where Consumers are Most Concerned About Voice Clones15.8%None of the aboveOtherSportsMedi

193、a/disinformationInsuranceHealthcareHome UtilitiesGamesPolitics/governmentBanking4.4%6.2%17.1%28.3%48.3%50.1%54.5%54.9%67.5%More than two-thirds of U.S.adults expressed concern about the risks posed by deepfakes and voice clones related to“Banking.”That was followed by the tight grouping of“Politics/

194、Government,”“Media/Disinformation,”“Healthcare,”and“Insurance,”that tallied responses ranging from 48%to 55%.Other categories were well behind,ranging from“Other”at 6.2%to“Home Utilities”at 28.3%.However,among the most notable findings is that only 4.4%of consumers could not identify a single indust

195、ry or domain that caused them concern,which means over 95%could.This figure accounts for consumers that did not know the terms deepfake or voice clone before learning the definition near the end of the survey.*For this question and those that follow,the survey participants were provided with term de

196、finitions.Earlier questions were posed without definitions.PAGE 61 VOICEBOT.AI-All Rights Reserved 2023Consumers Are Most Concerned About Banking RiskDEEPFAKE AND VOICE CLONE CONCERN BY DOMAINSource:Voicebot 2023Media/DisinformationHealthcareBankingInsurancePolitics/GovernmentNegativeNeutral+Dont Kn

197、owPositive$0-$50k63.5%49.8%51.6%47.4%45.1%$50k-$125k69.1%59.0%55.9%51.7%52.2%$125k+71.4%56.7%57.2%51.0%45.4%Income Impacts Domain ConcernTop Domains Where Consumers are Most ConcernedAbout Voice Clones and Deepfakes by Income U.S.adults across every income category were most likely to cite“Banking”a

198、s an area of concern related to deepfakes and voice clones.The frequency of citation rose with income level.Consumers with income below$50,000 were the least likely to express concern in every category,though the rate of citation for“Insurance”was nearly equal to the$125,000+income group.Consumers i

199、n the middle income group that ranged from$50,000-$125,000 were the most likely to cite“Politics/Government”and“Insurance”by meaningful margins.It also ranked highest for“Healthcare”but was less than 1%higher than the upper income group.PAGE 62 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE

200、CLONE CONCERN BY DOMAINSource:Voicebot 2023PAGE 63 VOICEBOT.AI-All Rights Reserved 2023Women Express More Concern in Direct Impact DomainsDEEPFAKE AND VOICE CLONE CONCERN BY DOMAINDomains Where Consumers are Most ConcernedAbout Voice Clones and Deepfakes by GenderIm not sureNoYesMedia/Disinformation

201、HealthcareBankingInsurancePolitics/GovernmentNegativeNeutral+Dont KnowPositive62.4%56.1%53.9%42.8%42.6%72.1%53.7%55.0%56.7%53.5%MenWomen Women are more concerned than men about the potential for negative impacts of deepfakes and voice clones in“Banking,”“Healthcare,”“Insurance,”and“Media/Disinformat

202、ion.”Men are more likely to express concern over“Politics/Government.”However,there is more to this story.While the gulf between men and women in“Politics/Government”and“Media/Disinformation”is within a couple percentage points,the variance for“Banking,”“Healthcare,”and“Insurance”are 9.7%,13.9%,and

203、10.9%,respectively.The services with a more immediate personal impact are more likely to be cited by women as areas of concern.Source:Voicebot 2023Consumer Expectations Differ for Banks and MediaPAGE 64 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE CLONE CONCERN BY DOMAIN Over 40%of consume

204、rs have high confidence that banks,insurance companies,and healthcare providers have taken steps to protect customers from deepfake and voice clone fraud.That same confidence falls to 34%for the news media and about 31%for social media.It is unclear if companies have taken steps to justify this conf

205、idence.Consider the inverse analysis which suggests that 30%have low confidence that banks,insurance,and healthcare have taken steps to protect their customers.That means bank customers have a confidence ratio of+11%,with about 30%unsure.In short,consumers have relatively high confidence that banks,

206、insurers,and healthcare providers are taking proactive steps to protect them.By contrast,consumers have lower expectations for news media and social media.The confidence ratio is-5%and-14%,respectively.Consumers are not optimistic that“media”organizations are proactively addressing this problem.Cons

207、umer Confidence in Deepfake and Voice CloneProtection for Key At-Risk IndustriesSocialMedia16.3%14.9%14.1%23.2%10.5%8.3%12.7%Extremely LowVery LowSomewhat LowNeutral or UnsureSomewhat HighVery HighExtremely High6.1%9.9%13.9%29.5%16.7%9.2%14.9%Banks,Insurance&Healthcare11.6%11.4%16.6%26.2%12.6%8.2%13

208、.4%NewsMediaSource:Voicebot 2023 While 40%of all consumers have high confidence that banks,insurance companies,and healthcare providers have taken steps to protect customers from deepfake and voice clone fraud,the figure rises to 50%for consumers with income higher than$125,000.Interestingly,the top

209、 income group is also the most skeptical that protections are in place,followed closely by the middle income group in the$50,000-$125,000 range.The means that the top income cohort has a deepfake protection confidence ratio of+16.8%,but that figure is just 6.7%for the middle income cohort which has

210、similar skepticism but far less optimism.PAGE 65 VOICEBOT.AI-All Rights Reserved 2023High Expectations of Banks,Insurance,and HealthcareDEEPFAKE AND VOICE CLONE CONCERN BY DOMAINConsumer Confidence in Deepfake and Voice Clone Readinessof Banks,Insurance,and Healthcare Providers by IncomePerceived De

211、epfake and Voice Clone Readiness ofBanks,Insurance,and Healthcare Providers by IncomeExtremely LowVery LowSomewhat LowNeutral or UnsureSomewhat HighVery HighExtremely High9.9%13.8%29.5%16.7%9.2%6.1%14.9%7.9%13.1%35.1%16.8%10.3%5.8%6.9%11.0%12.0%13.2%29.1%14.1%10.4%14.3%5.7%10.8%16.8%16.8%13.1%8.0%28

212、.9%$0-$50k$50k-$125k$125k+AllSource:Voicebot 2023Split Views About News Media Readiness Discussion about how deepfakes will impact political campaigns and government policy through the spread of misinformation in traditional media channels is on the rise.Consumers have less confidence that news medi

213、a organizations are taking proactive steps to combat the risk than they do for financial services and healthcare.This sentiment shows a news media deepfake protection confidence score of-5.4%across all U.S.adults.However,there is an interesting pattern among income cohorts.The highest income consume

214、rs have a+9.0%confidence score,while the middle and lowest income groups report-4.9%and-6.2%.Higher income is correlated with more confidence while middle and lower income are associated with more skepticism.Consumer Confidence in Deepfake and Voice Clone Readiness of News Media by IncomePerceived D

215、eepfake and Voice Clone Readiness of News Media by IncomeExtremely LowVery LowSomewhat LowNeutral or UnsureSomewhat HighVery HighExtremely High11.4%11.6%16.6%26.2%12.6%8.2%13.4%10.1%13.0%11.4%13.3%35.2%11.4%9.2%9.3%12.3%15.4%24.8%12.7%8.9%12.9%5.7%9.8%12.6%15.2%15.7%12.4%28.6%$0-$50k$50k-$125k$125k+

216、AllPAGE 66 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE CLONE CONCERN BY DOMAINSource:Voicebot 2023 Social media follows the income sentiment patterns of the other industries,but consumers express far more skepticism across each income cohort.The highest earners are,once again,the most pos

217、itive with a+1%deepfake protection confidence score.Another way to interpret the data is this group is equally optimistic and pessimistic.The most skeptical cohort is the middle income group with a-19.0%score,followed closely by-15.0%for the lower income group.That led to an aggregate confidence sco

218、re of-13.8%.Of course,social media is also where most consumers indicate they have encountered deepfakes and voice clones.This means they have first-hand experience with how the platforms police the technology.Skepticism About Social Media ReadinessConsumer Confidence in Deepfake and Voice CloneRead

219、iness of Social Media by IncomePerceived Deepfake and Voice CloneReadiness of Social Media by IncomeExtremely LowVery LowSomewhat LowNeutral or UnsureSomewhat HighVery HighExtremely High14.7%13.3%14.1%30.9%10.1%9.6%7.4%$0-$50k18.7%15.7%14.9%20.5%8.9%8.5%12.9%$50k-$125k7.7%13.1%17.8%10.6%16.0%9.5%25.

220、3%$125k+All14.9%16.3%14.1%23.2%10.5%8.3%12.7%PAGE 67 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE CLONE CONCERN BY DOMAINSource:Voicebot 2023All$0-$50k$50-$125k$125K+11.0%-5.4%-13.8%-15.0%-19.0%1.0%11.3%-4.9%6.7%-5.4%16.8%9.0%Banking,Insurance,&HealthcareNewsMediaSocialMediaHigh Income Opt

221、imists and Middle Income SkepticsDeepfake Protection Consumer Confidence Scoreby Industry and Income The summary table makes clear that the highest income consumers are the most optimistic about industries taking steps to protect their customers from risks posed by deepfakes and voice clones.Moreove

222、r,the data show a recurring theme that the middle cohort of$50,000-$125,000 annual income are the most skeptical.This means they are the most likely to express low confidence that these industries are being proactive.The area where the optimism outweighs the pessimism is in banking,insurance,and hea

223、lthcare.The lower income group falls in between the middle and upper income cohorts for each category.PAGE 68 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE CLONE CONCERN BY DOMAIN*The deepfake protection confidence score is calculated by subtracting the low confidence sentiment responses fr

224、om the high confidence.Source:Voicebot 2023Men Are More Confident Proactive Measures Are in PlacePAGE 69 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE CLONE CONCERN BY DOMAINPerceived Deepfake and Voice Clone Readiness of Banks,Insurance,and Healthcare Providers by Gender Men are far more c

225、onfident than women that banks,insurance companies,and healthcare providers have taken proactive steps to protect customers from deepfake and voice clone risks.Nearly half of men have at least“somewhat high”confidence that protections are in place.Those responses delivered a deepfake protection conf

226、idence score of+19.9%among men,while the same figure was-1.5%for women.This is a striking contrast.Less than 33%of women expressed positive confidence.This suggests women are more skeptical that these organizations are prepared for the risks,while men have relatively high readiness expectations.Sour

227、ce:Voicebot 2023Perceived Deepfake and Voice Clone Readiness ofBanks,Insurance,and Healthcare Providers by GenderExtremely LowVery LowSomewhat LowNeutral or UnsureSomewhat HighVery HighExtremely High9.9%13.8%29.5%16.7%9.2%6.1%14.9%9.6%12.0%25.5%15.4%11.2%5.7%7.3%20.6%10.7%16.4%32.7%14.0%8.2%10.7%Men

228、WomenAllPerceived Deepfake and Voice Clone Readiness of News Media by Gender Men express marginal confidence that the news media has taken proactive steps to automatically identify deepfakes and voice clones in their content,with a deepfake protection confidence score of just+2.8%.Of course,this is

229、far more favorable than the score registered by women of-10.3%.The rising concern about how deepfakes and voice clones may impact the spread of disinformation and reduce trust in news seems to resonate more with women,with over 40%expressing low confidence in the industrys readiness.This may be part

230、icularly relevant to U.S.television news where women are a majority of the audience according to Pew Research.PAGE 70 VOICEBOT.AI-All Rights Reserved 2023Men Are Less Confident in News Media ProtectionsDEEPFAKE AND VOICE CLONE CONCERN BY DOMAINSource:Voicebot 2023Extremely LowVery LowSomewhat LowNeu

231、tral or UnsureSomewhat HighVery HighExtremely High11.4%11.6%16.6%26.2%12.6%8.2%13.4%10.1%11.9%14.3%24.5%11.7%8.7%18.7%13.1%11.9%15.6%29.2%12.5%7.6%10.1%MenWomenAllWomen Believe Social Media is Not Ready for Deepfakes PAGE 71 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE CLONE CONCERN BY DOM

232、AIN While women express skepticism about deepfake and voice clone protections across industries,they are least confident in the readiness of social media platforms.Their deepfake protection confidence score is-24.7%,far more negative than the tepid skepticism from men at-4.4%.A 2022 Reuters Institut

233、e study found that only 26%of U.S.adults trust news organizations overall,a three percentage-point decline in one year and down from 34%in 2018.Moreover,“only 41%now say they trust the news they themselves use.”That is down from 50%in 2019.News media organizations are already facing a skeptical publ

234、ic and audience.The rise of deepfakes may cause a further erosion of trust.Perceived Deepfake and Voice CloneReadiness of Social Media by GenderSource:Voicebot 2023Extremely LowVery LowSomewhat LowNeutral or UnsureSomewhat HighVery HighExtremely High14.9%16.3%14.1%23.2%10.5%8.3%12.7%12.5%16.5%12.5%2

235、1.3%10.4%9.7%17.1%16.4%18.3%14.7%25.9%8.5%7.3%9.0%MenWomenAllDeepfake Protection Consumer Confidence Scoreby Gender and IncomeAllMenWomen11.0%-5.4%-13.8%-4.4%-24.7%19.9%2.8%-1.5%-10.3%Banking,Insurance,&HealthcareNewsMediaSocialMedia As you can see from the summary table,women are not confident that

236、 any of the industries we spotlighted in this report are taking proactive steps to protect consumers from deepfakes.While the sentiment is nearly neutral for banking,insurance,and healthcare,that is not the type of result most organizations aspire to.They would ideally like to have a strongly positi

237、ve sentiment among their customers.Social media has a particular challenge ahead to gain consumer trust regarding deepfake and voice clone protection.However,the data show that women in every group are skeptical and some high profile incidents of unchecked deepfakes could bring these numbers lower.P

238、AGE 72 VOICEBOT.AI-All Rights Reserved 2023Women Are Skeptical About Deepfake ReadinessDEEPFAKE AND VOICE CLONE CONCERN BY DOMAINSource:Voicebot 2023There is Broad Concern About Workplace ImpactPAGE 73 VOICEBOT.AI-All Rights Reserved 2023DEEPFAKE AND VOICE CLONE CONCERN BY DOMAINWorkplace Deepfake a

239、nd Voice Clone ConcernWorkplace Deepfake and Voice Clone Concerns15.1%Im notsure20.1%No64.8%Yes Nearly two-thirds of U.S.adults expressed concern about the inappropriate use of deepfakes or voice clones in the workplace.The expression of concern was higher among men than women by about 3.7 percentag

240、e points.The survey was not designed to clarify the source of the concern.However,several types of concerns may exist ranging from fraud attacks against the company or its customers,fraud attacks targeting a particular company employee,or inappropriate use of the technology that involves impersonati

241、ng or embarrassing a co-worker.This result is considerably lower than the more than 90%who expressed some concern about the technologies more generally.Source:Voicebot 2023Workplace Deepfake and Voice Clone Concerns by incomeWorkplace Deepfake and Voice Clone Concerns by income$50k-$125k$125k+All$0-

242、$50k64.8%20.1%15.1%57.0%24.2%18.8%66.0%20.6%13.4%75.0%13.4%11.6%YesNoIm not sure Similar to other findings,the concern about voice clones and deepfakes in the workplace rises with income levels.Only 57%of the sub-$50,000 income cohort expressed concern about the technologies impact in the workplace

243、compared with 66%in the$50,000-$125,000 range and 75%earning more than$125,000.These results pose an interesting question for enterprise organizations.While some are contemplating what to do about protecting the company and its customers from the threat of fraud,they may be overlooking employee conc

244、erns that involve internal use cases that disproportionately impact white collar workers.PAGE 74 VOICEBOT.AI-All Rights Reserved 2023Employee Concerns May Highlight Overlooked IssuesDEEPFAKE AND VOICE CLONE CONCERN BY DOMAIN*For this question survey participants were provided with term definitions.S

245、ource:Voicebot 2023Additional ResourcesFinancial Fraud DeepfakeU.S.Sen.Blumenthal Voice CloneSimon Cowell DeepfakeThe Daily Generative AI,LLM,and Synthetic Media NewsletterREAD NOWThe Latest Deepfake&Voice Clone Newsand InformationREAD NOWExposing the Truth About Zero-Day Deepfake AttacksLEARN MOREF

246、ortify Your Business Against DeepfakesWith PindropLEARN MOREGAIN WeeklyGenerative AI News ShowWATCH NOWPAGE 76More Deepfake ExamplesPAGE 76ResourcesLatestDeepfake NewsChart Index-Deepfake and Voice Clone Consumer Sentiment ReportPAGE 77Chart TitlePageDeepfake Technology Categories5Deepfake Solution

247、Complexity Map6Deepfake and Voice Clone Awareness Among Consumers7,53Deepfake and Voice Clone Concern Among Consumers7,56Domains Where Consumers are Most Concerned About Voice Clones8,61Consumer Confidence in Deepfake and Voice Clone Protection for Key At-Risk Industries9,64Deepfake and Voice Clone

248、Sentiment10Deepfake Consumer Encounters by Channel11,59Deepfake Awareness Among Consumers24Deepfake Awareness by Gender25Deepfake Awareness Among Select Media Consumers26Deepfake Consumer Encounters by Channel27Consumer Definitions of Deepfake Technology28Deepfake Sentiment Among Consumers29Deepfake

249、 Sentiment Among Media Consumers30What Consumers Like About Deepfakes31Deepfake Concern Level Among Consumers32Deepfake Sentiment Among Consumers with High Concern About Personal Risks33Consumers That Know Someone Who Has Had a Deepfake Made of Them34Percent of Deepfakes Consumers Say Were Made with

250、 Persons Consent35Deepfake Concern Level for Consumers That Know Someone That Has a Deepfake36Voice Clone Awareness Among Consumers38Voice Clone Awareness by Gender39Chart TitlePageVoice Clone Awareness Among Select Media Consumers40Deepfake Consumer Encounters by Channel41Consumer Definitions of Vo

251、ice Clone Technology42Voice Clone Sentiment Among Consumers43Voice Clone Sentiment Among Media Consumers44What Consumers Like About Voice Clones45Voice Clone Concern Level Among Consumers46Voice Clone Sentiment Among Consumers With High Concern About Per-sonal Risks47Consumers That Know Someone Who

252、Has Had a Voice Clone Made of Them48Percent of Voice Clones Consumers Say Were Made With Persons Con-sent49Voice Clone Concern Level for Consumers That Know Someone That Has a Voice Clone50Consumer Interest in Voice Authentication 51Interest in Voice Authentication Among Those Very Concerned About V

253、oice Clones51Deepfake and Voice Clone Awareness by Age54Deepfake and Voice Clone Awareness by Income55Deepfake&Voice Clone Concern Level Among Consumers by Income57Deepfake Sentiment Among Consumers58Top Domains Where Consumers are Most Concerned About Voice Clones and Deepfakes by Income62Domains W

254、here Consumers are Most Concerned About Voice Clones and Deepfakes by Gender63Consumer Confidence in Deepfake and Voice Clone Readiness of Banks,Insurance,and Healthcare Providers by Income65 VOICEBOT.AI-All Rights Reserved 2023Chart Index-Deepfake and Voice Clone Consumer Sentiment ReportPAGE 78Cha

255、rt TitlePageConsumer Confidence in Deepfake and Voice Clone Readiness of News Media by Income66Consumer Confidence in Deepfake and Voice Clone Readiness of Social Media by Income67Deepfake Protection Consumer Confidence Score by Industry and Income68Perceived Deepfake and Voice Clone Readiness of Ba

256、nks,Insurance,and Healthcare Providers by Gender69Perceived Deepfake and Voice Clone Readiness of News Media by Gender70Perceived Deepfake and Voice Clone Readiness of Social Media by Gender71Deepfake Protection Consumer Confidence Score by Gender and Income72Workplace Deepfake and Voice Clone Concern73Workplace Deepfake and Voice Clone Concerns by income74 VOICEBOT.AI-All Rights Reserved 2023Report AuthorsTo request custom AI industry research contact:infovoicebot.aiBret KinsellaResearch Director bretvoicebot.aiAndrew HerndonResearch,Design&Layout Editor