RN
The check came back empty, so start there.
Verifying a ranking like this one means opening ten profiles and reading a number off each. Attempting exactly that on one of the accounts involved returned an HTTP 402 and no page at all. Profile pages on X are not readable without an account, which means the figures underpinning any "most followed" list — including the ones this article used to carry — cannot be confirmed by a reader, cannot be updated, and cannot be audited by anyone who does not already have a logged-in session.
That is a fatal problem for a countdown, and it is worth being blunt about rather than papering over with a number that looks authoritative because it has a decimal point in it.
The metric was weak before it became unreadable
Set the access problem aside and assume perfect visibility. A follower count still would not measure what a list like this implies it measures.
Follower counts accumulate and do not decay. An account opened in 2011 carries every person who ever clicked follow, including those who have not opened the app in six years, those who followed for one viral post, and those whose accounts are dormant. Nothing subtracts. So the number is substantially a measure of tenure — how long the account has existed and how many news cycles it has been adjacent to — rather than of how many people see a given post today.
That is why these lists have a persistent bias toward performers with long careers and one widely-covered moment, and it is why they change so little year to year while actual attention moves around constantly. Ranking by a cumulative counter produces a leaderboard of seniority wearing the costume of a popularity chart.
What actually changed, and it is not the ordering
In early June 2024, X formalised an adult content policy. The written rule now says users may create, distribute and consume sexual material provided it is consensually produced and distributed and prominently labelled, and it frames sexual expression as a legitimate form of artistic expression. The policy explicitly extends to AI-generated images and video. Content promoting exploitation, non-consent, objectification or harm to minors stays prohibited, and adult material is withheld from minors and from adults who have opted not to see it.
Reporting at the time noted this largely codified what the platform had tolerated in practice under its previous name. But writing it down changed the situation materially: it made X the one large general-purpose social network with an explicit, published permission for adult content rather than an unwritten tolerance that could evaporate in a policy update.
For performers, that is the single most consequential fact about the platform, and it is worth more than any ranking. An unwritten tolerance cannot be planned around. A published rule can be — and can be pointed at when an account is actioned.
The labelling requirement quietly breaks the leaderboard
There is a consequence built into that policy that nobody puts in a countdown.
Labelled adult content is restricted from users who have opted out and from minors, and it does not circulate on the same terms as unlabelled material. So compliance with the policy costs reach by design. An account that labels correctly is deliberately reducing its own distribution; an account that does not is in breach.
Which means "most followed" and "most seen" have been pulled further apart by the very policy that made the content permissible. Two accounts with identical follower counts can have wildly different actual audiences depending on how much of their output is labelled. A ranking that ignores this is comparing numbers that no longer describe the same thing.
And then the audience became a question of geography
From 25 July 2025, Ofcom's requirement for highly effective age assurance applied to services hosting pornographic content in the UK. Self-declared age and payment methods that do not confirm an adult account are explicitly insufficient. Non-compliance exposes a service to fines of up to £18 million or 10% of global turnover, and to orders aimed at disrupting its business.
The practical effect for a reader is that whether you can see a given account's adult posts now depends on where you are and what you have verified, not only on whether you follow it. A single global ranking implies a single global audience. There is not one.
What a ranking would need before it was worth publishing
If someone wants to build this list properly, here is the specification, and it is demanding enough to explain why nobody does:
- A readable source. Counts observed directly, dated, and re-observable by the reader. Not a figure copied from another list that copied it from a third.
- A flow measure, not a stock one. Posts-per-week and replies-per-post over a fixed window, which decay when an account goes quiet. Cumulative followers do not.
- Consistent labelling treatment. Comparing accounts that label their adult content against accounts that do not is comparing different products.
- A stated jurisdiction. Since age assurance gates visibility regionally, "most followed" needs a "where".
- Verified identity. Impersonation and handle reuse are ordinary occurrences; a ranking entry that cannot be tied to the performer through a channel they control is a ranking of a username.
Meet those five and you have something defensible. Meet none of them and you have a countdown.
Where that leaves you
If you came here for the ten names, the honest position is that this site cannot verify them, and a list of unverifiable numbers dressed as a ranking is worse than nothing because it looks like a finding.
What is worth knowing instead: the platform's rules on this were written down in June 2024 and can be read, the same rules make labelled content deliberately less visible, and since mid-2025 whether you see it at all is partly a question of your jurisdiction. Those three facts will still be true next year, which is more than the follower counts managed.
The directory's X account listings and its performer databases are the practical starting points, and the databases are the better one — they exist to connect a performer to the channels they actually run, which is the step a leaderboard skips.