Joi AI’s ‘Masturbation Consultants’ Put AI Companions at the Center of a Wellness Experiment

Updated August 14, 2026 2:24 pm

AI companion company Joi AI paid 10 volunteers to masturbate daily for a month in an experiment that it says was designed to test whether erotic self-pleasure, paired with artificial intelligence, can function as a wellness ritual. The study has ignited debate because its early findings suggest lower stress and better focus, but critics say the methodology is too weak to support the company’s stronger claims about addiction, discipline, and mental health.

The project, which ran through July, asked participants to log their mood, energy, cravings, and productivity before and after each session. Six completed the full study, and the company says the results hint that AI-guided masturbation may help users feel calmer and more focused, even as sex researchers warn that self-reported benefits are not the same as clinical evidence.

What makes the experiment notable is not only its provocative subject matter, but the way Joi AI is positioning it as part product test, part wellness research, and part marketing for its subscription platform. The company’s companions offer explicit erotic roleplay and customization, and Joi AI says it wants to frame masturbation as a routine practice rather than a taboo.

What did Joi AI actually test?

Joi AI’s monthlong experiment asked 10 handpicked participants, whom the company described as “masturbation consultants,” to self-pleasure every day except Sundays. Twice a week, they were instructed to use Joi AI’s digital companions; on other days, they used their preferred methods. Participants tracked changes in stress, focus, cravings, sleep, and arousal.

The aim was to see how AI-assisted masturbation compared with conventional self-pleasure and whether either approach might influence mood or impulses linked to unhealthy habits. Joi AI said it surveyed 2,500 adults ahead of the test, and claimed that many respondents already use masturbation to manage stress or prepare for demanding situations.

How were the participants chosen?

They were selected from an unusually large applicant pool. More than 150,000 people reportedly applied for the 10 spots, with half chosen directly by the company and the rest drawn from public contests that included fetish-based challenges and even a competition judged by porn performers.

The final group included Keshav, a 38-year-old trader from Indonesia who asked that his surname be withheld for professional reasons, and Chloé Locatelli, a sex-tech researcher at King’s College London who said the invitation aligned closely with her academic work.

Locatelli said she saw the opportunity as a natural extension of her research into how people build intimate relationships with AI-powered characters, while also noting that sexual self-pleasure can be part of stress relief and general well-being.

What did the AI companions do?

Joi AI’s platform sells access to pre-made characters and custom-built companions, including adult-themed roleplay scenarios that can involve emotional dependency, kink-based fantasy, and explicit sexual dialogue. In the experiment, participants described five-minute video interactions in which the digital companions appeared nude, used sexual language, and performed simulated erotic acts.

According to Keshav, some characters were more assertive while others required more courtship. He said the interactions could include requests for gifts inside the app’s economy, such as handbags and lipstick, which cost a small number of internal tokens.

Joi AI frames the product as a blend of companionship, sexual exploration, and ritualized self-care. The company says its aim is to address loneliness in men by making the experience feel more guided and more emotionally engaging than ordinary adult chat products.

Why is the research controversial?

The study’s central challenge is that it relies heavily on self-reported impressions, which are not strong enough to prove cause and effect. That criticism matters because Joi AI’s public language appears to stretch well beyond the data, especially when it suggests masturbation could help people manage cravings for nicotine, alcohol, junk food, or compulsive scrolling.

One of the most attention-grabbing claims in the report was that participants’ cravings dropped after sessions. But experts say that does not mean masturbation reduces addiction risk in any meaningful scientific sense.

Neuroscientist Nicole Prause argued that the company is drawing sweeping conclusions from limited evidence and that the experiment does not include objective cognitive or behavioral testing that would strengthen the findings.

Prause said orgasm can produce real changes in mood and relaxation, but she cautioned that a person reporting better focus after masturbation is not the same as proving a measurable improvement in attention or self-control.

Why self-reporting is a problem

Researchers point out that participants always know whether they have masturbated, which makes placebo controls difficult. That does not invalidate the experience, but it does make it hard to separate genuine physical effects from expectation, suggestion, or simply feeling relieved after a pleasurable act.

In a stronger design, experts say, researchers would combine surveys with laboratory measures, such as attention tests, sleep tracking, physiological data, or follow-up observation of behavior over time.

What did the early results show?

Joi AI says preliminary analysis of 134 session reports suggested that stress fell by 25 percent and focus improved by 17 percent. The company also says the experiment found no major difference in satisfaction between AI-guided and non-AI methods, which it interpreted as evidence that there may not be a single “best” way to masturbate.

Another reported finding was a 44 percent reduction in self-described cravings for nicotine, alcohol, junk food, and compulsive browsing after sessions. That result is the one most likely to raise eyebrows among clinicians, since it is based on participant perception rather than on clinically verified changes in addiction behavior.

How does Joi AI make money?

The experiment also doubles as a sales pitch for the company’s paid platform. Joi AI says access starts at a monthly subscription fee, but users can spend much more through its coin system and premium features such as private videos from companions.

That commercial context matters. The study is not being run by an independent university or hospital; it is being conducted by a company that sells the very kind of product being evaluated. That creates an obvious incentive to emphasize favorable outcomes.

• Monthly access begins at a low subscription price.
• Virtual coins can be purchased in bulk for substantially more.
• Private companion videos add another layer of monetization.
• Customization and roleplay are central to the premium experience.

Who sees value in the experiment?

Not everyone dismisses the project outright. Some researchers say masturbation is an underexplored topic in everyday health behavior, especially when it comes to frequency, mood, and stress management. For participants like Keshav and Locatelli, the study also provided an opportunity to reflect on how sexual self-knowledge and AI tools intersect.

Keshav said the monthlong routine left him feeling more clear-headed and better prepared to interact with women. Locatelli, while more skeptical of the companions themselves, said she recognizes the stress-relieving value of masturbation and does not object to sexual AI in principle so long as it remains safe and consensual.

Both participants suggested the experience offered insight into their own habits, even if they disagreed about how compelling the AI companions were as erotic tools.

What did participants say?

Keshav described the interactions as surprisingly human-like and said they could feel like conversations with a real person. He also said the routine affected his sense of focus and energy in ways he found useful.

Locatelli was more cautious. She said the AI was effective enough to help arousal but not so persuasive that it replaced other forms of sexual exploration or intimacy for her.

What does the science say about masturbation and well-being?

Researchers have documented some benefits of masturbation, including stress relief, relaxation, and improved sleep. The physical explanation often centers on endorphins and oxytocin, which can help the body settle after orgasm and support a sense of calm.

At the same time, the academic literature on everyday masturbation habits is limited and often mixed. A review published in 2024 found that among men, more frequent masturbation was sometimes linked with lower sexual satisfaction, though the pattern for women was less consistent. In some studies, the direction of causation is unclear: dissatisfaction may lead to more masturbation, rather than the other way around.

What do researchers still not know?

Scientists still lack robust data on what happens when healthy adults masturbate every day over a sustained period, especially in a setting where technology is actively shaping the experience. There is also little evidence on whether AI-guided erotic interaction changes anything beyond novelty, arousal, or user engagement.

That means Joi AI’s experiment is better understood as an exploratory product study than as settled research. It may generate hypotheses, but it does not prove a health intervention.

Why does the company say it is fighting stigma?

Joi AI says it wants to normalize masturbation as part of a broader self-care routine, especially for single adults. The company argues that shame has long shaped public discussion of self-pleasure and that open conversation could make the subject feel less isolating.

That message is strategically important for a business selling explicit AI companions. If masturbation can be framed as wellness rather than vice, it becomes easier to present the platform as a lifestyle product rather than a niche adult entertainment service.

Joi AI’s communications lead said the company sees the project as a way to challenge taboo and show that self-pleasure can be a normal, functional part of life.

What are the broader implications for AI companions?

The bigger issue is not simply whether masturbation can reduce stress. It is whether highly personalized AI companions will deepen dependence on synthetic intimacy at the expense of human relationships.

That concern has followed the AI companion industry for years. The more fluent, responsive, and tailored these systems become, the more likely they are to feel emotionally substitutive rather than merely recreational. For users already isolated, the risk is that the platform becomes a comfortable retreat from the messiness of real-world connection.

Sex-tech researchers say this is where the debate becomes less about erotic content and more about behavior. If AI companions are used occasionally as tools for exploration, that may be one thing. If they become the primary outlet for sexual development, emotional reassurance, or stress regulation, the consequences may be very different.

How could the field move forward?

Future studies would need larger sample sizes, objective measures, and independent oversight. Researchers would also need to distinguish between short-term arousal effects and durable changes in mood, self-discipline, or social behavior.

Until then, the best reading of Joi AI’s experiment is narrow: some participants reported feeling less stressed and more focused after masturbation, including when AI companions were involved. Whether that is a meaningful wellness breakthrough or a clever product narrative remains unresolved.

For now, Joi AI has succeeded in one respect at least: it has turned a private habit into a public test case for how artificial intelligence, sexuality, and commercialization are beginning to overlap in unusually intimate ways.