In just three months, TikTok in the US removed 781.5 million fake followers from accounts and blocked 475.8 million inflated likes. These figures come from the second quarterly report by TikTok USDS Joint Venture, the company managing the US TikTok app since January 2026. The report, released on October 20, covers the period from April 1 to June 30, 2026.
For short-form video creators, this isn't just abstract data—it's a price list of risks. Everything sold under the banner of a “growth service” is being removed by the platform in droves and is considered a violation of its rules.
What TikTok Removed This Quarter
The platform tackles two main areas: combating fake activity (followers, likes, spam accounts) and standard video moderation. Here's a look at the former, compared to the previous quarter.
| Metric | Q1 (Jan–Mar) | Q2 (Apr–Jun) |
|---|---|---|
| Fake followers removed | 1.23 billion | 781.5 million |
| Follow attempts blocked | 302.7 million | 403.2 million |
| Fake likes prevented | 95.9 million | 475.8 million |
| Spam account creations blocked | 30.5 million | 64.9 million |
| Videos removed with spam accounts | 9.6 million | 18.8 million |
| Comments removed with spam accounts | 4.9 million | 5.9 million |
The platform also reported removing over 14 million accounts identified as fake during the quarter. The announcement itself rounds this figure to 14.6 million, while the report's tables for the same period list it as 14.7 million.
Fewer Followers Removed, But Five Times More Likes
The most interesting part of the table isn't a record-breaking number, but the trend. The number of fake followers removed in Q2 was significantly lower than in Q1: 781.5 million versus 1.23 billion. However, the number of blocked likes skyrocketed fivefold, and blocked follow attempts increased by a third.
The platform doesn't explain this shift directly, but the logic is clear from other figures: the number of spam accounts blocked at creation doubled (from 30.5 million to 64.9 million). The more bots that are prevented from entering the system in the first place, the fewer fake followers need to be pruned later.
Networks: 30 Commercial and 4 Political
Beyond automated mass removals, the platform also conducts targeted investigations. During the quarter, it dismantled 30 coordinated networks in the US that were artificially inflating engagement, involving 76,000 accounts and their 4.8 million posts. According to the report, a typical member of such a network isn't an ideologue but someone capitalizing on trending topics to drive views for financial gain.
Separately, there were four covert influence operations: networks of accounts that coordinated to push political narratives to US and global audiences. The platform lists their tactics: fabricated personas, location spoofing, pseudo-news accounts, re-uploading others' content, and AI-generated characters and videos. Another 817 accounts were removed for attempting to reinstate previously dismantled networks.
How the Platform Removes Videos
The second part of the report covers standard content moderation. The key takeaway for creators here is that the platform rarely removes content based on user reports; it finds almost everything on its own.
| Metric | Q1 | Q2 |
|---|---|---|
| Total videos removed | 8.41 million | 12.23 million |
| Removed automatically, without a human moderator | 7.45 million | 11.01 million |
| Videos restored after appeal | 571,800 | 587,300 |
| Removed before any user reports | 98.8% | 99.2% |
| Removed within the first 24 hours of posting | 90.9% | 89.4% |
| Accounts removed for being fake | 12.6 million | 14.7 million |
Two more stats for Q2: 82.2% of removed videos were taken down before receiving a single view, and 97.7% of user reports were processed in under two hours.
Meanwhile, the share of removed videos relative to all published content is tiny—just 0.4%. This means the mass purges primarily affect bot farms, not everyday creators.
A Closer Look at Fake Engagement and AI
Within the 'Integrity and Authenticity' category, the platform provides a breakdown by specific policy violations. For Q2:
- Deceptive behavior and fake engagement — 43,894 videos. Of these, 99.1% were removed before any reports, but only 60.3% were caught before the first view and 71.1% within the first 24 hours. This is the slowest category for automated systems to catch: fake engagement is harder to identify from the video itself and is more visible through account behavior.
- Altered media and AI-generated content — 646,835 videos, up from 377,218 in the previous quarter. That's a nearly twofold increase. This isn't about banning AI, but about enforcing rules on labeling and passing off generated content as real.
What This Means for Creators
- Don't buy followers and likes. The 781.5 million followers removed in a single quarter is exactly what happens to purchased engagement. The money is gone, the numbers roll back, and your account is left under suspicion.
- Check if you've been targeted by fake engagement. If your follower count suddenly drops, it's likely a purge, not a glitch. This can happen even if you didn't buy anything yourself, due to old contractors, “gifted” fakes, or acquiring an account with a history of inflated numbers.
- Re-uploading is also a risk. Posting someone else's video without your own creative input falls under the same authenticity policy as fake personas. If you work with clips, your contribution must be clear: editing, an intro, subtitles, or commentary.
- Label your AI content. Videos with AI content are the fastest-growing category for removals. The platform's rules require you to label generated content, especially if it depicts realistic-looking people or events.
- Removals are fast and usually automated. 99.2% of violating videos are removed without a user report, and 82.2% before their first view. There's no point in hoping your video will slip through the cracks. However, the appeals process works: 587,300 videos were restored last quarter.
If your videos suddenly stop getting views, it's not always because they were removed—sometimes it's a reach limitation. We broke down how to tell the difference in our article on the TikTok shadowban. For the platform's latest content and account requirements, check out our breakdown of the new TikTok rules.
What This Means for Brands
For advertisers, this report is a strong argument for paying for views, not followers. Followers can be removed retroactively: a creator might have 300,000 today and 180,000 after the next purge, long after your budget is spent.
What you should do:
- Look at the views on their last 10–20 videos, not their follower count.
- Be wary if an account has many followers but few views and comments.
- Pay for results that can't be bought in bulk, like completed views.
- Don't demand “guaranteed” reach from creators, as these guarantees are often met using fake engagement.
We explored how pay-per-view works in practice and how much of that payment goes to intermediaries in our article on how much clippers get paid.
What We Still Don't Know
The platform didn't disclose everything, and a few things remain unclear from the report:
- The consequences for accounts that receive fake engagement. The report has numbers on removed followers but no data on sanctions against the recipient accounts, such as how many faced reach limitations, warnings, or bans.
- How many of the 781.5 million fakes came from commercial “growth services.” The platform doesn't distinguish between purchased followers and those distributed by botnets on their own.
- There's no country-by-country breakdown. The entire report is only about the US app managed by TikTok USDS Joint Venture, which covers 200 million users and 8.5 million businesses in the US. It contains no specific figures for Russia, Belarus, or Kazakhstan.
It's important to understand this last point correctly: while the numbers are from the US, TikTok's authenticity rules and anti-fraud systems are global. For a creator outside the US, this report isn't just news about America—it's a data-backed look at the mechanics at play on their own account.
Frequently Asked Questions
What happens if you buy fake followers on TikTok?
The inflated followers get removed—781.5 million were removed in the US alone in one quarter. Buying engagement falls under the policy on deceptive behavior and fake engagement. Nearly 44,000 videos were removed for this violation in Q2, and offending accounts can face restrictions up to and including a ban.
Why did my TikTok followers suddenly drop?
It's most likely a scheduled purge of fake accounts, not a bug. Even if you didn't buy any, a drop is still possible: some of your followers may have been bots that were mass-following accounts.
Is inflating views and likes a violation?
Yes, the platform considers artificial likes, follows, views, and other methods of artificial amplification to be fake engagement. In one quarter in the US, it blocked 475.8 million fake likes and 403.2 million follow attempts.
Can I get removed followers back?
No. You can appeal the removal of a video or an account—587,300 videos were restored last quarter. Removed fake followers cannot be recovered.
Does this report apply to countries like Russia and Kazakhstan?
Not directly, as it only contains data for the US app. However, the platform's authenticity rules and enforcement mechanisms are global, so the takeaways apply to any account.
In Short
In one quarter, TikTok in the US removed 781.5 million fake followers, blocked 475.8 million fake likes and 403.2 million follow attempts, prevented the creation of nearly 65 million spam accounts, and dismantled 30 fake engagement networks and 4 covert influence operations. 99.2% of violating videos were removed before anyone reported them.
Buying fake engagement is a waste of money. A working alternative is taking on brand tasks that pay for real views. On Prime Oracles, brands post tasks in the “Content Rewards” section. A creator can accept a suitable task by clicking “Accept Task” and get paid for the views their video generates. A brand can launch a campaign by clicking “Create Campaign.”
Source: The TikTok USDS Joint Venture report “Protecting authenticity on TikTok” from October 20, 2026, and its accompanying Community Guidelines and Platform Security Reports.
