OnlyFans turns ten: the platform's empowerment pitch is colliding with evidence of an extractive layer beneath it

Ten years after launch, OnlyFans is the rare British-headquartered platform whose cultural footprint has outrun its headquarters in Camden, north London. The brand — sold in 2018 to a Ukrainian-born founder, later the subject of a contested ownership transition, and now majority-owned by a US-listed parent — has been sold to the public, simultaneously and contradictorily, as a bastion of creator autonomy and as a familiar content site with familiar risks. On 18 June 2026 a Guardian editorial marked the anniversary by redirecting attention to a less flattering layer of the story: the agents, "chatters," and middlemen who, the paper says, take up to 50% of some women's earnings. The argument is uncomfortable, and it deserves more than ritual denial.

The central tension is not new, but the anniversary has made it newly legible. OnlyFans markets the work as entrepreneurial — a woman, alone, with a smartphone and a 20% platform fee, building a direct line to paying subscribers. The Guardian's editorial undercuts that image with a portrait of an intermediary layer that looks rather like the talent management structures the platform claims to have made obsolete. If agents are taking half, the headline autonomy number (the 80% the creator keeps, net of OnlyFans' own cut) is no longer the relevant figure. The relevant figure is whatever reaches the creator after the agency stack has been paid.

A "creator economy" with a manager class

The pattern the editorial describes is unglamorous. The paper's reporting points to recruitment on social media, WhatsApp-led onboarding, and tiered contracts that route the agency's share off-platform through separate payment rails. From the platform's vantage point, none of this has to exist. OnlyFans provides payments, hosting, and moderation; the rest is, formally, between the creator and whoever she signs with. From the creator's vantage point, the maths is the maths. A signed account with three full-time "chatters" producing 16-hour-shift direct-message engagement does not behave like an individual business; it behaves like a small agency with a single principal and asymmetric information.

This is the part of the story that does not fit the empowerment pitch. Empowerment as a commercial brand requires the subject of the pitch to be the relevant unit of analysis. Once the agency is reintroduced — particularly an agency that takes a larger cut than OnlyFans itself — the unit shifts back to the firm, and the firm looks, in several respects, like the management agencies the platform was supposed to render unnecessary.

Why the UK framing matters

The Guardian is explicit that it is writing to a UK audience and addressing UK Members of Parliament. Ofcom already has a residual role in the regulation of some online harms, and the Online Safety Act has expanded the duties of platforms operating into the UK market. None of those regimes was drafted with this specific market in mind. Adult content has historically been treated as a category the regulator would rather not name, and the result is a patchwork: payment processors carry much of the de facto governing weight, because they are the ones with the contractual leverage to make platforms behave.

The editorial's proposal — that MPs look at the agents, the chatters, and the recruitment pipelines — is, on the evidence it cites, a sensible one. The agency layer is a labour-market phenomenon, not a hosting-market phenomenon. As a labour question, it has natural homes in HMRC's enforcement of employment status, in the Department for Business and Trade's interest in unfair-contract terms, and in any parliamentary committee inclined to ask whether the term "self-employed" still describes the work in question.

The counter-narrative: autonomy, agency, and the right to choose badly

OnlyFans' defenders, including the platform itself in past statements, will argue that any voluntary contract between an adult creator and a manager is the business of those two parties, and that the presence of a 50% agency cut does not, on its own, prove exploitation. Some creators will say they tried management, found it unsuitable, and walked away with the platform fee as the only overhead. Others will say management bought them time, removed the worst of the direct-message labour, and was worth the share. Both accounts can be true at once.

The harder question is what the platform is for, if not for protecting the version of the deal in which the creator is the principal. The empowerment pitch implicitly promises that direct-to-consumer monetisation eliminates the agent. If the agent reappears, with renewed leverage, then either the pitch was always incomplete, or the platform's role has to be redefined — for example, by surfacing verified information about agencies, by capping disclosed agency cuts in creator-facing interfaces, or by requiring escrow-style payment structures that make the agency take visible at the point of contract. The platform has, to date, not moved in this direction with any public urgency.

Stakes, and what remains unsettled

If the Guardian's reporting holds up, the consequence is a re-rating of the platform's social claim. A host that bears 20% and a manager that takes up to 50% leave the creator with a third of gross, before tax and before the cost of production, lighting, and the labour of being a person who performs this work. That is a defensible business for someone with no other options; it is a less comfortable basis on which to assert that the platform has "empowered" its creators as a class.

What remains genuinely unsettled is the denominator. The Guardian's editorial cites a 50% figure; it does not — in the form available to this publication — claim that this is the modal or median cut. The agency model, if it exists at scale, is heterogeneous: some agents will charge more, some less, some none at all. The existence of a problematic 50% structure is not, on its own, evidence that 50% is typical. The reasonable inference is that the structure exists and that its existence deserves parliamentary attention, as the editorial argues. The further inference — that the platform's empowerment pitch is hollow across the board — would require data the editorial does not, on its face, supply.

What is clearer is the political shape of the next twelve months. A UK government that has shown willingness to legislate on platform liability, on payment-processor responsibility, and on the umbrella category of online harm now has, in the editorial, a roadmap for a narrow but defensible intervention: treat the agent layer as the labour question it appears to be, and let the hosting question rest where it currently sits. Whether ministers pick the roadmap up is a question the editorial, fairly, leaves to the MPs it is addressing.

*Desk note: this article reads the Guardian editorial as the wire signal it is — a 700-word framing piece aimed at Westminster, written on the platform's tenth anniversary, citing a single 50% figure. Monexus's contribution is to read that editorial against the platform's longstanding empowerment pitch and to mark, plainly, where the editorial's case is strongest (the structural reappearance of an agency class) and where it is thinnest (the lack of a denominator for the 50% figure).