A class action, a 1956 psychology paper and a new academic study all point at the same thing — and it isn’t the content.
One of the plaintiffs in a class action against OnlyFans said he had signed up mainly to have friendly conversations with creators and to show them photographs of his cooking.
That detail was reported as comedy when the case surfaced. It is closer to the rule than the exception, and there is now peer-reviewed work explaining why.
A chapter published this January in the Palgrave volume The Shaping of the Parasocial Self examines exactly this question. Johanna L. Degen of the University of Flensburg analyses subscription platforms from both the consumer and performer side, and reaches a conclusion that cuts against almost everything written about the subject in the popular press: the drivers are the need for intimacy, for contact that feels continuous and reciprocal, and for personal validation. Sexuality and pornography, she writes, are mechanisms that only secondarily intersect with that.
Not incidental. Secondary. The sexual content is real and it matters, but on this reading it is the medium rather than the motive.
The concept is seventy years old
Degen’s framework rests on an idea introduced in 1956, when Donald Horton and R. Richard Wohl published a paper in Psychiatry on what they called parasocial interaction: the sense of intimacy at a distance that audiences develop toward performers who cannot see them.
They were writing about television hosts. The mechanism they described — a one-directional bond that the audience experiences as a relationship — needed no modification to explain a subscription platform seventy years later. What changed is that the bond now has a price, a payment rail and a direct message inbox.
That is the whole innovation. Not the intimacy, which is old. The billing.
The word doing the work is “supposedly”
Degen’s phrasing is careful. The contact is supposedly reciprocal.
That qualifier is where the business model lives. A subscriber who believed they were reading generic broadcast content would not pay for it, because free equivalents are infinite and one search away. A subscriber who experiences the exchange as directed at them personally will pay, and will keep paying, and — as the platform’s own revenue mix shows — will pay again for each unlocked message.
So the product is not the photograph. The product is the impression that someone specific is on the other end.
This also explains something that otherwise looks irrational. Free tube sites carry more content, in higher volume, at zero cost. They lose comprehensively to a paid platform on this one dimension, and they cannot fix it, because the thing being sold is not available in a library.
Then the question went to court
In July 2024, a group of subscribers filed a class action alleging that the person answering their messages was frequently not the creator at all but a professional “chatter” employed by a management agency, working from scripts designed to increase spending.
Court reporting on a related 2025 complaint describes how the plaintiffs came to suspect it: the message volume did not add up. One creator’s account had hundreds of thousands of subscribers, and no single person could sustain that correspondence.
The litigation has not gone the plaintiffs’ way. A judge dismissed the case in December 2025 with leave to amend, reasoning that the relevant possibilities were disclosed in writing. In May 2026 a California federal judge largely granted motions to dismiss the amended claims, throwing out most of what remained against the platform operators while allowing one narrow claim to proceed against the management agencies over the sharing of personal information.
Read alongside the research, the outcome is more interesting than a win would have been.
The court’s position is that the terms disclosed what might happen. The research position is that the product only functions if the subscriber does not experience it that way. Both can be correct simultaneously, and the gap between them is not a bug in the model. It is the model.
What this means if you are the one paying
If the thing you are buying is a relationship with one particular person, then which person matters far more than it would on a content site.
On a tube site the unit is the clip and the creator is nearly interchangeable. Here the unit is the creator, and switching costs are emotional rather than financial. That makes the initial choice unusually consequential — and the platform offers almost no help with it. There is no browse function, no comparison view, no way to see who works in which niche at what price before committing.
What exists is external. Independent indexes compile public profile data into something you can filter, and this search tool is one of them, which is a strange thing to have to say about a business whose entire value proposition rests on picking the right person.
Two things worth knowing before you do. Volume is a signal: an account with a very large subscriber base cannot plausibly be answering everything personally, whatever the tone of the replies suggests. And many creators disclose agency involvement somewhere in their terms, which is exactly the disclosure the court found sufficient.
The part nobody has solved
Degen’s account has an uncomfortable implication that the litigation stumbled into.
If what is being purchased is the experience of reciprocity rather than reciprocity itself, then the human on the other end is, strictly speaking, an implementation detail. Agencies worked that out years ago. The platform’s 2026 rules now require a human somewhere in the chat loop, which is a tacit acknowledgement that the question had become live.
Seventy years after Horton and Wohl described intimacy at a distance, the open question is no longer whether people will form attachments to someone who cannot see them. It is how far away that someone is allowed to be.