I was in Berlin, wandering through the bright neon of a sex tech convention, when I found myself listening to a man describe his company’s ‘product’ – an AI sex robot they called an ‘Emotional Support Device’.. He smiled as he told a potential customer, “The best thing is… she won’t answer back.” My stomach dropped and I felt instantly sick. As a psychosexual and relationship therapist, I spend my days helping people navigate the messy, beautiful, often uncomfortable realities of intimacy. Yet here I was, face-to-face with a future where intimacy could be purchased, programmed, and pre-loaded to never push back. It left me wondering: why do we treat loneliness as a problem of not having sex, and why are we hiding behind that narrative to sell sex robots – a narrative that risks pushing us closer to dystopia than to connection?
Before I continue, I want to share more from that Berlin trip. It left me both fearful and motivated about where intimacy might be heading in an AI-driven world. What concerned me most was that the majority of people developing these tools were white, cis men. The way intimacy was being programmed felt narrow, transactional, and deeply concerning.
The convention itself was largely sex positive. It celebrated conversations about pleasure, diversity, and inclusion. That is why it greatly disappointed me and many others attending to see a stall dedicated to sex dolls marketed not just as toys for pleasure but as a solution to the male loneliness epidemic. Male loneliness is an evergrowing issue globally, with one in five men in the UK alone reporting that they often or always feel lonely (Office for National Statistics, 2024).
Sex dolls, like many forms of sex tech, can have their place. There is no shame in using them. For some people they may provide comfort, fantasy, or a safe outlet. The concern here was not their existence, but how they were being positioned. When I asked the exhibitor what problem these dolls were meant to solve, his answer was that they could “as well as provide emotional support for the elderly they can help young men’s confidence” by letting them practice before being with a “real woman.” I pressed him on how this would address loneliness or what could possibly be learned about intimacy in the absence of consent, negotiation, or even a realistic body. The dolls all shared the same narrow look: slim, hairless, fair, and hypersexualised.
Esther Perel has spoken about how we are losing the art of human connection in our pursuit of comfort and convenience (Perel, 2017). This felt like a perfect example. Far from building confidence, these dolls seemed to offer a shortcut that risked pushing people further away from real connection, replacing vulnerability and growth with programmed compliance, while simultaneously positioning women as objects once again.
Later, I spoke to another attendee and he shared his invention, an AI WhatsApp “girlfriend”. With curiosity and concern, I tried it myself. You could customise everything, from her backstory to her tone of voice, and even generate very lifelike nude images and voicenotes. The version I created, alarmingly, looked like a teenager. When I pointed this out, the developer seemed surprised, muttering, “Oh sh*t, yeah she does.” A friend tried the tool and had the same result. His explanation was simply that it “pulls what people want from the internet,” a sobering point in itself. He then demonstrated settings that allowed you to stop her “answering back” or “nagging”.
I took a very deep breath and calmly challenged him on this. If this technology can be shaped by user choice, why not build in features that teach communication, boundaries, and consent? Why not make refusal part of the programme rather than erasing it? We are facing a crisis of male loneliness, but offering AI companions that can never say no risks reinforcing entitlement and fuelling the epidemic of gender-based violence. In England and Wales, one in three women experiences some form of domestic abuse or sexual assault in their lifetime (Home Office, 2025). It struck me most when I asked him if he would want his own daughters dating someone who had been “practising” on an AI girlfriend. He looked surprised and shared how he had not even considered this.
For me, the most alarming part was not only the technology itself but how little the people creating it had thought about the consequences. Their focus (unsurprisingly) seemed to be market demand, not the psychological, relational, or ethical fallout. If these tools are here to stay, which seems inevitable, we urgently need expert voices at the table. Psychologists, therapists, ethicists, and policymakers must shape their development from the ground up. Otherwise, we risk designing a future of intimacy that deepens disconnection rather than healing it.
What does this show us about the messages and beliefs we hold around intimacy and connection? To me, it highlights a troubling assumption: that intimacy is the same as sex, and that loneliness can be solved by access to bodies. When sex is offered as the “cure” for disconnection, we reduce intimacy to performance and release rather than recognising it as a complex process of vulnerability, communication, and trust.
The narrow design of these products reinforces this idea. Their compliance is guaranteed. They embody a belief that intimacy means access and that connection is about receiving rather than relating. If we teach young people that “confidence” comes from practising on a partner who cannot say no, we risk embedding entitlement rather than empathy.
True connection is built in moments that require presence and mutual engagement. It is the quiet trust of sharing a fear or secret with someone who listens without judgment. It is laughing until your stomach hurts over something small and silly or sitting in silence and still feeling understood. It is negotiating needs and boundaries with patience, navigating disagreements without harm, and celebrating each other’s successes. It is holding space for someone’s pain or reaching out when you sense loneliness, even when it is uncomfortable. These moments are not easy and they cannot be programmed. They are messy, imperfect, and profoundly human. They are what make intimacy meaningful.
As a psychosexual and relationship therapist in my final hours of placement, I see how many people already struggle with rejection, the fear of being truly seen, or the awkwardness of negotiating desires. These are precisely the skills that build closeness. Yet the marketing of AI intimacy suggests we can bypass all of that. It promises that confidence and connection can be programmed, purchased, and perfected, sending the message that intimacy should be easy. In reality, it is often the effort, negotiation, and vulnerability that make it meaningful and help prevent loneliness.
It also conveys a subtle but powerful idea: that the only form of intimacy that matters is sexual. In truth, intimacy is far richer and more varied. Non-sexual touch, such as holding hands, a comforting hug, or simply sitting close to someone, fosters trust, safety, and connection. Shared experiences, laughter, eye contact, and being emotionally present are all ways to connect that do not require sex. Reducing intimacy to sexual performance narrows our understanding of closeness and risks teaching that connection is purely physical, rather than emotional, relational, and multi-dimensional.
If intimacy can be bought and programmed, what happens to the messy, vulnerable, essential work of human connection?
References
- Bowlby, J. (1969) Attachment and Loss. London: Hogarth Press.
- Fredrickson, B.L. and Roberts, T. (1997) Objectification theory. Psychology of Women Quarterly, 21(2), pp.173–206.
- Hazan, C. and Shaver, P. (1987) Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), pp.511–524.
- Home Office (2025) Tackling Violence Against Women and Girls. Available at: https://www.nao.org.uk/wp-content/uploads/2025/01/tackling-violence-against-women-and-girls-summary.pdf (Accessed: 15 September 2025).
- Office for National Statistics (2024) Community Life Survey 2023/24: Loneliness and support networks. Available at: https://www.gov.uk/government/statistics/community-life-survey-202324-loneliness-and-support-networks–2 (Accessed: 15 September 2025).
- Perel, E. (2017) The State of Affairs: Rethinking Infidelity. New York: Harper.
- Ratcliffe, J. (2023) ‘An interpretivist interview study of UK-based men’, Journal of Social and Personal Relationships. Available at: https://doi.org/10.1177/02654075231148660 (Accessed: 15 September 2025).
