TikTok fake accounts are not a small background problem. In its latest U.S. enforcement update, the platform says it removed 14.6 million fake accounts during the second quarter of 2026 and stopped tens of millions more spam accounts from being created.
For creators and marketers, the more revealing number is 781.4 million fake followers removed. It is a reminder that follower counts can change because a platform cleans up artificial activity, not just because audiences lose interest. The reported removals say something important about the scale of enforcement, but they do not tell advertisers how much of any particular creator’s following is genuine.
Table of contents
Jump to each section:
- What TikTok’s second-quarter numbers actually cover
- Why fake followers complicate creator metrics
- What marketers should check before signing a creator
- The limits of a platform cleanup number
What TikTok’s second-quarter numbers actually cover
The figures cover April 1 through June 30, 2026, and come from the second U.S. Platform Security and Community Guidelines Reports issued by TikTok USDS Joint Venture. The company announced the second-quarter reports on October 20, following its first U.S. reporting round in July.
According to Social Media Today’s breakdown of the data, TikTok reported four separate account-integrity measures:
- 14.6 million fake accounts removed: Accounts TikTok says it took down after identifying them as inauthentic.
- 64.8 million spam accounts prevented: Sign-up attempts blocked before the accounts were created.
- 781.4 million fake followers removed: Inauthentic follow relationships removed from the platform.
- 403.2 million fake follow requests prevented: Attempts to generate artificial follower activity that were blocked.
Those numbers should not be added together as though they represent unique people or unique accounts. Removed accounts, blocked registrations, deleted follower relationships and blocked follow requests describe different actions.
TikTok’s U.S. team also reported disrupting 30 covert influence networks, with more than 76,000 associated accounts and 4.8 million posts removed. That is a separate enforcement category and should not be confused with the fake-follower totals.
Why fake followers complicate creator metrics
Brands still routinely begin a creator shortlist with audience size. TikTok’s figures show why that is an incomplete starting point. If a platform removes fake follower relationships at this scale, a creator’s displayed follower count alone cannot verify the quality or reachability of the audience behind it.
A follower cleanup does not necessarily mean a creator purchased those followers. Fake accounts may follow legitimate profiles without the account owner’s involvement. Nor does a falling follower count, by itself, establish that a creator has violated platform rules.
The reverse is equally important: a large audience is not evidence of fraud. The second-quarter figures are aggregated U.S. enforcement data, not an audit of individual creators. TikTok did not publish enough information in the reported figures to calculate what percentage of U.S. followers were fake or to determine how representative the removals are of a typical influencer account.
For creator campaigns, this makes the quality of actual content engagement more useful than a static follower total. Brands should pay attention to whether recent videos consistently reach the audience they want, whether comments relate to the content, and whether campaign results match the audience profile promised in the brief.
What marketers should check before signing a creator
The report gives marketers a practical reason to review how they vet and measure creator partnerships. Three checks are especially useful before a campaign goes live.
- Look at a sequence of posts, not one viral hit. Review a creator’s recent content over a consistent window, including typical views, shares and meaningful comments. Outlier videos can distort an average.
- Compare audience fit with campaign goals. Check geography, language and relevant interests using platform or creator analytics where available. A big audience in the wrong market may be less valuable than a smaller audience with genuine category interest.
- Put business outcomes beside engagement. Use tracked links, first-party conversions, qualified visits, codes or other agreed campaign measurements. For awareness campaigns, document the measurement limits instead of treating follower growth as a sales result.
Brands should also agree in advance how reporting handles deleted, blocked or otherwise invalid engagement. A campaign recap that uses only the creator’s headline follower count may miss changes that occur between selection, publication and reporting.
None of these checks can guarantee that an audience is entirely authentic. They do, however, make a partnership decision less dependent on one number that a platform can revise through enforcement.
The limits of a platform cleanup number
The most tempting interpretation of TikTok’s figures is that more removals automatically mean more spam, or that the moderation system has become more effective. Neither conclusion follows from the totals alone. Higher enforcement volumes can reflect more abuse, improved detection, policy changes or some combination of those factors.
WeRSM’s October 28 report also points to the breadth of TikTok’s moderation effort. Regulated goods and services represented the largest reported video-removal category, followed by body exposure and sexualized behavior. The regulated-goods category includes services that sell artificial likes, views and followers, alongside other prohibited or restricted activity.
There is a distinction between enforcement activity and measured platform integrity. The published counts demonstrate how much activity TikTok says it intercepted. They do not independently establish how many fake accounts remain or whether any given campaign’s performance is free of manipulation.
For marketers, the takeaway is not to abandon TikTok or discount every creator metric. It is to make audience quality, content performance and attributable outcomes part of the same conversation. The platform can remove millions of bad signals, while brands still need their own evidence that the signals left behind are worth paying for.