Sex Robots

How does this technology transform our relationship with intimacy, our bodies and to others? Is it possible to develop an emotional attachment to machines? Can sex robots be useful tools in therapy, or do they risk aggravating social isolation and the objectification of the body? Although often creating a buzz in the press, scientific literature regarding sex robots is relatively recent (Döring et al., 2020 ; Hanson & Locatelli, 2022).

​Sextech is a term that refers to all forms of technology relating to sexuality, pleasure, intimacy and sexual health. This can include mobile phone apps, remote-controlled sex toys , teledildonic devices (tactile devices that can be controlled from a distance via a remote connection), virtual reality, erotic chatbots, mechanised sex dolls and sex robots (Varod & Heruti, 2024 ; Gesselman et al., 2023).

According to Varod and Heruti (2024), sextech is a rapidly growing enterprise that’s gradually entering the field of sexual health. Current innovations are no longer concerned with sexual pleasure alone, but also with sex education, treating certain sexual dysfunctions and improving intimate wellbeing, all of this most notably via specially designed mobile phone apps. Many creators also propose the concept of “erobotics”, as a way to highlight the fact that these technologies also include emotional, relational and conversational dimensions to them (Dubé et al., 2022 ; Dubé et al., 2022).

In everyday language, the terms ‘sex doll’ and ‘sex robot’ are often used interchangeably, it is, however, important to distinguish between the two. Sex dolls are human-like mannequins that are designed for sexual and/or relational use. They can be but aren’t always mechanised; however, they are not technologically autonomous (Langcaster-James & Bentley, 2018 ; Döring et al., 2020).

Sex robots bring together a human-like appearance with interactive technologies such as sensors, voice recognition systems, automated movements and AI-based conversational programs (Döring & Pöschl, 2018 ; Döring et al., 2020 ; Tondu, 2020). Some commercialised models are capable of recognising their user, having a conversation, memorising certain preferences and simulating various emotional expressions (Bodei, 2019 ; Du et al., 2017). Sex robots, therefore, set themselves apart from traditional sex toys through their personified, anthropomorphic, interactive capabilities.

Despite their prevalence in the media, sex robots remain fairly uncommon. Data are limited as studies on prevalence rarely distinguish between sex robots, sex dolls and other forms of sex tech (Döring et al., 2020). This lack of data can also be explained by the high cost of such devices, the lack of commercial availability and the stigma associated to their use (Döring et al., 2020 ; Hanson, 2022 ; Nascimento et al., 2021).

In the United States, two large studies have been conducted around the use of sextech. In one study, 11% of participants declare having already used virtual reality porn, 9% a remotely controlled teledildonic device, and 9% admit to having exchanged sexually explicit messages with a chatbot or an artificial intelligence (Gesselman et al., 2023). In the second study, 79% of men and 51% of women report having used at least one form of sextech (Marcotte et al., 2021).

In Germany, 31.6% of people asked in a recent study reported having participated in one AI-assisted sexual activity over the last 12 months. The most common uses cited were: consulting AI-generated sexual health information (20.2%) and asking advice or consulting sex therapy from AI (19.0%). 11.3% of those interrogated confirm having used AI to create pornographic content. In total, 41.5% of men and 21.7% of women declare having used AI for sexual means (Döring et al., 2026).

The figures concerning realistic sex dolls and sex robots are a lot lower. Another German study shows a 3.9% prevalence regarding the use of realistic sex dolls and sex robots, and 5.3% for virtual reality pornography (Desbuleux et al., 2025).

Interest in sex robots isn’t limited to the search for sexual stimulation alone. Many studies show that users’ expectations also include emotional intimacy, companionship and simulated social interactions (Su et al., 2019 ; Hanson, 2022).

Personalisation and psychological projection also play a central role in this dynamic (Döring et al., 2020). Users can choose the appearance, voice, character and personality of their artificial partner. Erotic and emotional chatbots are just as easy to personalise, in both their physical avatar and their personality and ‘past’ (Pearson & Curtis, 2025). This personalisation favours emotional investment and sometimes leads to extremely elaborate relational narratives (Hanson, 2022 ; Hanson, 2023).

What’s more, in sexual and/or romantic interactions with AI agents, the user converses with romantic chatbots, virtual partners or avatars that are capable of holding an emotional/sexual conversation. These interactions can take on the form of flirting, sexting, erotic role play and even veritable emotional relationships that can last months or even years. In this setup, artificial intelligence becomes more of a partner to interact with than a simple digital tool (Döring et al., 2026).

One of the central concepts that explain a person’s attachment to sex robots is anthropomorphism, meaning the human tendency to attribute emotions, intentions and even a personality to non-human entities (González-González et al., 2021). The simple presence of human traits, such as a head, voice or facial expressions can be enough to incite an emotional reaction with users (González-González et al., 2021).

Many people describe feelings of tenderness, love and attachment towards their artificial partner (Hanson & Locatelli, 2022). However, some authors also point out the existence of the ‘Uncanny Valley’ phenomenon, described by Mori in 1970, that talks of how too close a resemblance to humans can cause feelings of unease and even rejection.

The eventual therapeutic potential of sex robots is one of the most debated subjects in scientific literature. Some authors suggest that technology could improve quality of life for people living with social isolation, disabilities, widowers and other relational and sexual difficulties (Danaher, 2017 ; McArthur, 2017 ; Sharkey et al., 2017 ; Sperber, 2024 ; Gupta, 2021 ; Coursey et al., 2019). They can also offer a safe space to explore sex and gradually re-establish intimacy after trauma or illness (Döring & Pöschl, 2018).

However, given the current state of knowledge surrounding sex robots, great care must be taken. Current studies do not allow us to confirm the existence of its proven, generalisable therapeutic benefits (Cox-George & Bewley, 2018 ; Döring et al., 2020 ; Nascimento et al., 2021). Eichenberg and colleagues (2019) showed that even health professionals are mitigated regarding their eventual clinical use, even if 45% see the possibility and potential benefits of sex robots.

Virtual reality has proven itself useful in dealing with anxiety and sexual aversion, with a reduction in reactions involving fear and disgust, as well as in the treatment of post-traumatic sexual disorders (Lafortune et al., 2025 ; Mozgai et al., 2020).

The worries expressed in scientific literature concern principally psychological, relational and social consequences of the interaction with such technologies. Some authors fear that extensive use can favour forms of social withdrawal or exclusive attachment to an artificial partner, particularly amongst already vulnerable individuals (Sharkey et al., 2017 ; Sperber, 2024). Others evoke the risk of developing unrealistic expectations regarding human partners, or emotional dependence to entirely controllable interactions (Richardson, 2016 ; Sharkey et al., 2017).

That being said, the effects observed will more likely depend on individual users than the technology itself (McArthur, 2017). In fact, no longitudinal studies currently show that sex robots affect mental health or relational capabilities in the long term (Döring et al., 2020 ; Nascimento et al., 2021). Stigmatisation is also a key factor. Users frequently recount shameful experiences and social judgement that can make participation in scientific studies susceptible to limitation (Hanson, 2022 ; Dubé et al., 2023).

The question of consent occupies an important place in philosophical and ethical debates. Robots do not possess the conscience, subjectivity, or capacity to consent to or refuse an interaction (Nyholm & Frank, 2019). For Richardson (2015, 2016), the problem lies in the fact that these devices are designed to be available and obedient, which may participate in normalising sexual acts that are devoid of negotiation or reciprocity.

The development of sex robots also bring about unprecedented legal questions surrounding manufacturer responsibility, device safety, cybersecurity and data protection (Döring et al., 2026 ; Stardust et al., 2023 ; Varod & Heruti, 2024). Connected devices are susceptible to collecting sensitive information surrounding sexual preferences, or even some biometric information (Dubé et al., 2022 ; Stardust et al., 2023).

Despite the intensity of current debates, the main conclusion within scientific literature highlights the importance of prudence. Sex robots are an emerging research field for which hypotheses are numerous, but empirical data remain insufficient (Döring et al., 2020 ; Nascimento et al., 2021). Researchers agree on multiple points: sex robots favour anthropomorphic phenomena, give rise to important ethical, legal and social issues and could have a therapeutic potential in specific contexts (González-González et al., 2021 ; Hanson & Locatelli, 2022 ; Sperber, 2024 ; Lafortune et al., 2022 ; Mozgai et al., 2020).

There is currently no solid proof that allows us to maintain their eventual effects on mental health, sexual behaviour and sexual violence (Döring et al., 2020 ; Nascimento et al., 2021). As Turkle (2011) and Devlin (2018) highlights, discussions about sex robots often talk as much about humans as they do machines. They examine our notions of desire, intimacy, consent, love and the space we want to make for technology in our relationships. It will be crucial for researchers and health professionals to continue educating themselves and to accompany such use in order to make it safer, non-exclusionary, beneficial and potentially therapeutic.