Design, Use, and Effects of Sex Dolls and Sex Robots: Scoping Review

‎

Research Findings on Sex Robots

In the following sections, the main findings of previous sex robot research will be reported separately for the 6 groups of sex robot publications ().

Sex Robot Conceptualization and Theory

The largest group of sex robot publications (40/98, 41%) deals with the conceptualization of sex robots and of human–sex robot relationships. Within this group, 2 issues are predominant: the (non)inherent sexism of sex robots and the (non)humanness of sex robots.

Echoing the critical feminist conceptualization of female sex dolls, several publications on sex robots characterize the female sex robot as an inherently sexist object. The most cited author of this position is Kathleen Richardson [], who conceptualizes the female sex robot as a representative or surrogate of a sexually objectified woman, a female (forced) porn actor, a female (forced) prostitute, or a female sex slave. Following this conceptualization, the production and use of female sex robots is regarded as harmful for individual male users, their female partners, and society at large, as female sex robots symbolically reinvent and reaffirm the status of women as sex slaves []. However, this conceptualization operates more with metaphorical equations than established theories and is challenged by other publications as vague and unconvincing [,,]. Although existing sex robots might appear sexist, different designs are possible; therefore, sex robots are not inherently sexist, according to other authors [,,,,].

Further theoretical publications deal with the question of the humanness of the sex robot. Several publications stress that, by their definition, sexual interactions and intimate relationships are bidirectional and require a consenting human partner. Humanness—by the definition of these publications—implies sentience, first-person consciousness, and free will; none of these attributes can be ascribed to current sex robots. Consequently, the authors conclude that current sex robots are nonhuman pseudo persons. Accordingly, relations with robots are only pseudorelationships that inherently lack mutual concern for the welfare of each other [] and do not lead to personal or spiritual growth []. Following this conceptualization, there is no sexual interaction possible with a robot or between a human and a robot, only robot-enhanced solipsistic masturbation [].

There are also publications that focus on future advanced sex robots and their humanness. Several authors assume that, in the foreseeable future, sex robots could be produced that are sentient, self-conscious, and have a free will []. They might even have the legal status of citizens so that humans can legally marry them [,,]. Such advanced humanoid robots will be so human-like that they must be conceptualized as persons and relations with them as interpersonal relationships. Advanced sex robots with excellent social and sexual skills and perfect looks who enter relationships with humans out of their own free will could be very attractive for many people []. Advanced sex and love robots could bring more love to the world [], but they could, at the same time, devalue real humans []. However, the concept of an advanced sex robot that is almost indistinguishable from a real human paradoxically makes it seem pointless to build sex robots. If the advanced sex robot acts like a self-determined, willful human, if it consequently lies, cheats, criticizes, disregards, rejects, and leaves the human, what is the merit of creating it in the first place [,,,]?

Obviously, there is an inherent tension in the conceptualization of the degree and quality of humanness of interactive humanoid robots. A dumb robot is easy to control but lacks autonomous capabilities and sociability; hence, it cannot bring much additional value to traditional sex dolls. An intelligent advanced robot provides true sociability but lacks the manageability and obedience that we expect from a service technology.

A third conceptualization overcomes the divide between a sex robot as a mere masturbation aid and a sex robot as a quasihuman and stresses that successful sex robots can easily be imagined as purposefully designed to be nonhuman like regarding appearance, functionalities, and social role. Possibilities might be sex robots as synthetic animals [], as fantasy creatures, or as interfaces to other types of sexual entertainment technology (fantasy hardware) []. Sex robots could be cherished and desired by humans, particularly by digisexuals or technophilics, precisely because of their fundamental otherness [,]. As humans can bond with sex dolls, it is even easier for them to form meaningful emotional attachments with interactive sex robots [,,]. Instead of insisting that advanced sex robots be as human-like as possible to legitimize sexual interactions and emotional attachments with them, robots could also be accepted as nonhuman social agents, for example, to provide safe sex work [], alleviate social and sexual deprivation [], or allow for safe explorations of sexual fantasies [].

The different implications of a wider use of sex robots are addressed by many theoretical publications as unanswered questions [,], for example, regarding health [], social norms [], and religious beliefs [].

Ethics of Sex Robots

What is the right thing to do in view of the emergence of sex robots? The second largest group of sex robot publications (28/98, 29%; ) attempts to tackle this core question of sex robot ethics. Although some authors stick to metareflection and debate which ethical approach to use [], other authors provide answers of 3 different types:

• Sex robots should not be built and used at all. Starting from the assumption that human-human sexual and romantic relationships are most healthy and ethically superior to all human-robot pseudorelations and to the use of sexualized and sexist robotic objects, authors with different political [], philosophical [], feminist [,], theological [,], engineering [], and legal [] academic backgrounds reject further developments in this field. They call for bans and boycotts, stigmatization of, and abstinence from sex robots. Quite popular and often quoted in the media are the arguments of the earlier mentioned Kathleen Richardson, founder of the Campaign against sex robots, who compares sex robots with killer robots and with female sex slaves [,].
• Sex robots should be built in an ethical way to avoid harm to humans, especially vulnerable humans. Starting from the assumption that sex robots can be a good thing if they alleviate loneliness and/or sexual deprivation and contribute to the sexual and social well-being of individuals and couples, authors with different backgrounds encourage ethical design []. Publications are very diverse and often vague as to what exactly they expect from ethical sex robot design. One author explains that she wants sex robots designed in such a way that they do not get involved in acts of infidelity because they have learned the concept of heartbreak []. Other authors explicitly do not want sex robots to be love robots because they fear humans could be too easily manipulated by robots that fake romantic attachment [,]. Others want sex robots designed in a women-friendly [] and disability-inclusive [,] way or demand design that is more focused on consumer safety []. Some authors point out the many different questions for ethical design ranging from “Should the robot become active on its own and entice the partner to have sex?” to “How should the robot collect an evaluate patient data to better satisfy its partner’s sexual needs?” [,]. As some authors assume that the development of sex robots is driven by a profit-oriented uncaring industry [], there is a need for more involvement of ethically responsible entrepreneurs and designers from different backgrounds who aim to develop and market sex robots for sexual well-being, pleasure, fun, and play while taking into consideration the concerns and desires of diverse user and stakeholder groups. Some authors are very optimistic that sex robots will bring a lot of pleasure and happiness and are, therefore, ethically a good thing [], although some ethical issues are unresolved (eg, regarding robot prostitutes) []. Other authors stress that robots are a good thing only for a very small group of people who absolutely cannot find a human sex partner [].
• Sex robots should be built in an ethical way to avoid harm to robots, especially for advanced sentient robots. Starting from the assumption that future humanoid robots will be advanced to a very high degree of human likeness, according to several authors, their sexual and other citizen’s rights must be protected with a nonanthropocentric but robocentric ethic [,,,,]. For example, sex with an advanced sex robot should only be acceptable if the robot has given explicit consent []. Although some authors stress the relevance of a robocentric ethic for sex robots to protect them from anticipated harm and exploitation, other authors argue that sentient robots designed as sex robots could have a “good life” and experience pleasure and satisfaction from fulfilling their tasks [].

Empirical Studies on Sex Robot Use and Effects

The third group of sex robot publications contains empirical studies (12/98, 12%; ). Thus far, not a single empirical study has been published that deals with the small but presumably growing number of pioneer users of sex robots. All existing studies address nonusers and investigate their attitudes toward sex robots and their reactions to sex robot–related stimuli.

Most empirical studies (8/12, 67%) are small web-based surveys on sex robot acceptance using convenience samples from the United States (N=261: [], N=133: [], N=100: [], and N=198: []), Germany (N=263: []), Indonesia (N=380: []), and Malaysia (N=32: []). Their results show diverse rates of sex robot acceptance. For example, 40% of male and 17% of female respondents in the United States (mean age, 33 years) reported willingness to try out a sex robot [] in comparison with 16% of Indonesian respondents [] and 9% of Malaysian respondents []. Cultural background, male gender, positive attitudes toward robots in general, interest in manga and games, sensation seeking, and shyness appeared to be predictors of sex robot acceptance. Interestingly, sexual and relationship satisfaction did not predict sex robot acceptance [,]. However, because of the small nonrepresentative samples, the generalizability of existing sex robot acceptance data is very limited. Another problem is the varying operational definitions of sex robots given to respondents in the surveys. A Delphi survey explored the predictions of 20 social robot experts who were reluctant regarding sexual apps [], whereas 1 expert interview explored the sex robot predictions of the founder of sex doll and sex robot manufacturer Abyss Creations, Matthew McMullen [].

In addition to the survey and interview studies, 3 experimental studies were found (3/12, 25%). They investigated how heterosexual women experience their male partner’s imagined infidelity with a female robot vs a real woman [], at which body parts of female robots vs female humans, both represented in pictures, male and female gaze [], and how humans physically react when they touch different, including private, body parts of a robot that is not a sex robot []. Overall, these experiments show differences and similarities in humans’ sexuality-related reactions to humanoid robots and fellow humans. So far, no experimental study exists that uses an actual sex robot as the stimulus material.

Sex Robot Representations in Art and Media

The fourth group of sex robot publications concerns sex robot representations in art and media (8/98, 8%; ). In their selected and analyzed examples from the science fiction literature, some studies from humanities focus on fictional female sexbots that seem to embody male fantasies of the ideal woman but who, in the course of the action, become feminist robots striving for independence from their male human partner or creator by leaving or even killing him, for example, the robot Ava in the 2015 UK movie Ex Machina [,], the virtual Samantha in the 2013 US movie Her [], or the robotic wife in the 1981 Chinese story Conjugal Happiness in the Arms of Morpheus []. The famous US television series Star Trek Voyager presented the character Seven of Nine, a cybernetic organism and former Borg drone, who—although embodying traditional feminine beauty—challenged traditional ideas of gender and sexuality []. Other dystopian science fiction representations, selected and analyzed by the academic literature, illustrate the female sex robots’ sexual exploitation and victimization, for example, as porn actors in the 2009 US movie 2040 [] or as rape victims in the US television series Westworld []. One monograph critically analyzes posthuman utopias in sex robot representations [], and one editorial volume documents the Robot Love 2018 International Expo of the Niet Normaal Foundation in the Netherlands that brought together researchers and artists [].

A quantitative media content analysis examined the representation of human–sex robot relationships in 370 fictional and 340 nonfictional media examples []. The results of this study indicate that media representations of intimate human-robot relationships tend to portray the human partner as a man who is disadvantaged in interpersonal relationships. At the same time, media often portray the involved robot partner as a humanoid female sex robot. Although nonfictional media describe intimate human-robot relationships more often in sexual terms, fictional media focus more on emotional aspects, cohabitation, and even procreation between humans and robots. Overall, media representations of intimate human-robot relationships reveal stereotypical gender roles, heteronormativity, and a focus on sexual vs emotional intimacy [].

Legal Regulation of Child Sex Robots

The fifth group of sex robot publications covers child sex robots (6/98, 6%; ). All the 6 publications [,,,-] characterize child sex robots as harmful and unethical and call for a legal ban that is already in preparation or in effect in several countries (eg, the aforementioned CREEPER Act of 2017 in the US) []. In all, 2 publications speculate on the possible therapeutic uses of child sex robots. Although one of them assumes that their exploration would be too risky [], the other encourages their exploration only in certain, controlled circumstances under strict medical supervision and in accordance with guidelines issued by an ethics committee [].

Design of Sex Robots

Only 4 publications in the sex robot literature focus mainly on design (4/98, 4%; ): 1 on erotic voice output [], 1 on a mind-controlled neurodildo to be used separately or implemented in robots [], 1 on general design aspects based on results about sex doll use [], and 1 on feminist sex robot design in an analogy of initiatives for feminist pornography [].

Research Gaps in Sex Robots

Although the body of sex robot publications is 3 times as large as that of sex doll publications, empirical studies on sex robot use and effects are equally scarce (4 peer-reviewed papers in total). Fundamental questions regarding the sexual use of human-like full-body material artifacts that remained unanswered for sex dolls also remain unanswered for sex robots.

Research Gaps in Sex Robot Design

Just as with sex dolls, the question of how much fantasy, and which and whose fantasies should legitimately be implemented in sexual fantasy products to make them socially acceptable, harmless, and still sexually desirable and exciting, remains unanswered with sex robots as well. Although there is a lot of speculation on the possible therapeutic uses of sex robots to be found in public and academic debates, the literature fails to provide design guidelines for therapeutic sex robots informed by evidence from sex and relationship therapy and focused on specific problems (eg, on sexual shyness and anxiety, sexual dysfunctions, sexual trauma, paraphilias, and paraphilic disorders). Design studies for current sex robots hardly exist, and the literature predominantly speculates about imagined future sex robots. Instead of researchers, it was journalists who first dealt with the question, “What would sex robots for women look like?” [] and who let women and men draw and explain their ideal sex robots [].

Research Gaps in Sex Robot Use

Thus far, no empirical study has investigated experienced sex robot users or interactions of unexperienced participants with actual sex robots. Results from research on sex doll use are, therefore, the best available proxy for sex robot use. Regarding the potential market size and user population, there are also no data available that allow for sound predictions. One might speculate that sex robots could overcome some of the stigmatization of sex dolls as sex robots can be framed as cutting edge, high-tech products. Thus, their users might appear more modern, future oriented, and competent in comparison with traditional sex doll owners. Against this backdrop, one would expect more growth for the sex robot market than for the sex doll market, but data are needed. To further explore the sexual appeal of robots, insights from research on objectophilia [,] and technofetishism [] could be helpful.

Surprisingly, the sex robot literature falls short in conceptualizing and investigating interactions and relationships between humans and current sex robots in a psychologically nuanced way. Whereas the sex doll literature has already established that dolls easily trigger humans to build meaningful, caring, loving, long-term relationships with them, the sex robot literature often falls back on binary thinking. It categorizes the current sex robot as an inanimate object and mere masturbation aid without any sociability and is only willing to ascribe sociability to future imagined sex robots that are advanced to the point of indistinguishability from humans. Hence, the literature on sex robots often misses the key point that robots are more than mere masturbation aids due to anthropomorphization and that they are meaningful and possibly helpful precisely because they are not substitutes for real humans but are sociotechnical entities for parasocial use and play. Parasocial interactions and play with sexual fantasy products grant more degrees of freedom in sexual expression and allow to take a break from all of the norms, ethics, expectations, and responsibilities of human-human interactions.

Research Gaps in Sex Robot Effects

As sex robot users and use are completely unknown thus far, any claims about positive and/or negative effects are mere speculation. Although some authors are so convinced of their speculations on strong to catastrophic negative effects that they demand immediate boycotts and bans of sex robots, others urgently call for empirical research on sex robot effects. The idea that sex robots allow humans to indulge in interactive embodied sexual fantasies elicits strong projections of lust and fear. Most likely, empirical research will help us overcome exaggerated projections and understand the diversity and ambivalence of effects on different types of sex robot users.

Discussion

Main Results of the Review

In conclusion, the main results of the whole review are summarized, its limitations are indicated, and a roadmap for future research is drawn. The body of sex robot literature, with approximately 100 academic publications in total, is more than 3 times larger than that on sex doll literature, with approximately 30 publications (RQ1). However, only a handful of peer-reviewed empirical papers on both sex doll use and sex robot use are available thus far. No sex robot study exists that investigates people experienced in sex robot use and/or introduces actual sex dolls or sex robots as stimulus material. Regarding the first RQ, one must concede that sex dolls and sex robots, although attracting growing public and scholarly attention, are heavily under-researched. Both sex doll and sex robot research are fields characterized by disciplinary diversity, with notable participation from philosophy, humanities, and engineering, and a conspicuous lack of participation from sex researchers.

Sex doll and sex robot designs (RQ2) are often critically assessed in the literature, mainly because the bodies of women-like dolls and robots are usually designed in sexualized ways following and exaggerating traditional feminine beauty ideals. However, when understanding sex dolls and sex robots as sexual fantasy products, it makes sense that they do not imitate reality but cater to sexual fantasy. Often, it is exactly the point of sexual fantasies to be unrealistic. Thus far, the literature has not addressed the core question of how we could and should assess designs of sexual fantasy products such as sex dolls and sex robots, considering both social inequalities and vulnerabilities and the freedom of sexual fantasy and expression. Regarding future advanced sex robots, the literature presents various requirements for ethical design, which—at the current state of robot development—are very speculative. Systematic design studies that work with current and future users (eg, private sex doll owners, sex workers, and sex therapists) and address different use scenarios (eg, domestic, commercial, or therapeutic) are lacking.

Although previous research has provided some insights into the domestic long-term use of sex dolls (with or without parallel psychotherapy), no data have been collected thus far on the short-term interactions or long-term relationships between humans and sex robots. Thus, the best proxy for sex robot use and users today is the limited data on sex doll use and users (RQ3).

Considering the lack of empirical knowledge about sex doll users and sex robot users and use, it is obvious that the predictions of positive and negative effects found in the literature can only be speculative (RQ4). It is striking that authors still provide very strong and contradictory effect claims ranging from utopian visions of improved sexual satisfaction and overall happiness to dystopian visions of dehumanization, objectification, and isolation. Predictions of small and/or ambivalent effects might be more realistic but are seldom discussed in the academic literature thus far, which seems to mirror some of the hype and scandalization observable in public media discourses.

Limitations

This scoping review addressed sex doll and sex robot research as far as it is represented in the accessible literature published before August 2019. We were particularly careful to retrieve publications not only from the databases but also, in a systematic way, from all the included publications’ reference lists. Nevertheless, it must be taken into consideration that further studies that have not (yet) been published and/or were not (yet) accessible (eg, conference presentations, qualifications theses, and journal articles under peer review) could exist. However, we are confident that our systematic literature identification strategy covered previous research thoroughly enough, especially as this is the very first systematic review of the field.

To map previous research in a comprehensible and useful manner, we organized the body of literature by building distinct groups of publications according to their key topics and methodologies. We discussed data charting and synthesis within the team and checked everything in duplicate. However, some decisions might be questionable. There is an inherent tension between the aim of providing a clear and comprehensible structure, which requires a reduction in complexity, and the aim of doing justice to the individual publications, which requires a representation of their complexities. Due to space constraints, we were forced to reduce the complexity much more than we would have wished. Therefore, we encourage readers to consult the original publications whenever in doubt and apologize to fellow researchers in case they feel our review misrepresents their work.

Another inherent problem of a multidisciplinary review lies in the tension between the aim of doing justice to discipline-specific styles of knowledge production and communication and the aim of presenting existing knowledge in a consistent, readable, and generally understandable way. We deliberately simplified concepts and streamlined discipline-specific jargon to improve consistency. We provide a broad overview spanning from ancient Greek myths to contemporary web-based surveys, and spanning from the psychoanalyst’s office to the robotics lab. We agree with many authors we cite in this review that a deeper understanding of sex dolls and sex robots and their meanings for human sexuality can only be achieved through more interdisciplinary collaboration. We hope that our prioritization of disciplinary width over depth will inspire this collaboration. However, we are aware of the risks and limitations of simplification.

Roadmap for Future Research

Hopefully, the many and diverse research gaps pointed out in this review can serve as starting points for future research projects on sex dolls and sex robots, their design, use, and effects. To conclude, we suggest 4 selected, particularly urgent research strands:

1. Public debates about and media representations of sex dolls and sex robots, polarized and scandalizing as they are, attract much attention, shape public opinions, and influence research activity. They deserve more scholarly analysis and participation by the sex research community. This includes traditional mass media and social media. Mass media tend to assume dramatic positive or negative effects, while often completely ignoring the fact that sex dolls and sex robots, overall, could have only small and/or ambivalent effects. Social media, sometimes, offer more nuanced views, but expert statements and documentaries about sex robots on YouTube, for instance, are met by a noteworthy amount of misogynist comments that welcome female sex robots as substitutes for women. On Twitter and Instagram, we see sex dolls communicating to the public, their accounts steered by doll owners (eg, the earlier mentioned Davecat) and by doll manufacturers. These examples illustrate that we need to know more about media representations as they are an important element of the cultural context in which sex dolls and sex robots are developed, marketed, discussed, used, and investigated today.
2. Research on the sexual uses of human-like material artifacts such as sex dolls and sex robots needs to be advanced and connected to research on human-like digital artifacts such as chat bots, avatars, holograms, or immersive virtual reality pornography. After all, a sexual AI system trained by a particular user could be used on different technological and media platforms such as a full-body sex robot, an immersive virtual reality system, or a smartphone. Although the materiality of dolls and robots offers new possibilities in terms of embodied sexual fantasies (eg, physical presence, physical care, physical touch, and physical stimulation), it also creates boundaries (eg, through the body weight and difficult handling of the dolls and robots at home, limited mobility outdoors, high visibility, and risk of social stigmatization). For sexual fantasy products that aim to enhance their users’ sexual and social experiences, the right degree and mixture of materiality and virtuality is an open question for research and design.
3. Although sex robots have triggered the publication of many theoretical and ethical papers, we urgently need empirical data on actual sex doll and sex robot users and uses. Different study designs (nonexperimental and experimental, cross-sectional, and longitudinal) and data collection methods (qualitative interviews, focus group discussions, surveys, psychological tests, and physiological measurements) are suitable for research with actual sex doll and sex robot users. Instead of using only imagined or visually depicted artifacts as stimulus material, some of the real sex dolls and sex robots should be incorporated in empirical studies.
4. Despite the relatively large number of theoretical papers, the degree of theoretical elaboration of human–sex doll/sex robot relations and their consequences is not yet very high []. Commonly used theoretical concepts are objectification, gratification, and pseudorelationships. For a more thorough understanding, we suggest including theoretical concepts from the field of doll play and doll therapy and from the field of human interaction with media personas (parasocial interactions and relationships), with digital technologies (computers as social actors and media equation theory) and with social robots (uncanny valley concept and anthropomorphization) as well as from social psychology (social cognitive learning theory), clinical and developmental psychology (transitional objects, objectophilia, and robophilia), and sexuality research (sexual scripts theory and theories on sexual fantasies). It is not yet clear which theories from the different related research fields on dolls, robots, sexuality, gender relations, well-being, and health are best applicable to human–sex doll/robot relationships and if and how they can be combined to best explain the complex intimate engagements of humans with artifacts.

Conflicts of Interest

None declared.

‎

‎

Abbreviations

Edited by G Eysenbach; submitted 04.03.20; peer-reviewed by O Bendel; comments to author 25.03.20; revised version received 03.04.20; accepted 03.04.20; published 30.07.20

Copyright

©Nicola Döring, M Rohangis Mohseni, Roberto Walter. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 30.07.2020.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.