Skip to main content
Community Guidelines Enforcement

Community Guidelines Enforcement Report

January 1, 2023 – March 31, 2023
Published June 30, 2023

About this report

TikTok is a global entertainment platform fueled by the creativity of our diverse community. We strive to foster a fun and inclusive environment where people can create, find community, and be entertained. To maintain that environment, we take action upon content and accounts that violate our Community Guidelines or Terms of Service and regularly publish information about these actions to hold ourselves accountable to our community.

TikTok uses a combination of innovative technology and people to identify, review, and action content that violates our policies. This report provides quarterly insights into the volume and nature of content and accounts removed from our platform.

Analysis

TikTok’s Community Guidelines are designed to foster an entertaining experience that prioritizes safety, inclusion, and authenticity. Our policies apply to everyone and all content, and we strive to be consistent and equitable in our enforcement. This analysis provides additional context to supplement the data in this report.

Safety

In this report, we’re sharing new insights as we continue to bring transparency to how we operate and protect the platform. First, we’re providing insight into the language capabilities of TikTok’s moderators whose work not only contributes to a safer platform but also plays a vital role in enhancing our automated moderation systems. Second, we’re providing data on our response time to remove violative content reported by our community using our in-app and online reporting tools.

During this report period, we updated TikTok’s Community Guidelines. These guidelines went into effect in April 2023 and will be reflected in our next report.

Security

We remain vigilant in our efforts to safeguard the platform from adversarial threats, including the presence of inauthentic or fake accounts and engagement. These threats persistently probe and attack our systems, leading to occasional fluctuations in the reported metrics within these areas. Despite this, we remain steadfast in our commitment to promptly identify and remove any accounts, content, or activities that seek to artificially boost popularity on our platform.

Protecting our community is our priority. While TikTok is a global platform, we take a localized approach to regulatory compliance, collaborating with stakeholders to gain a deep understanding of local concerns and fulfill our regulatory commitments. In the U.S., we have established a dedicated division called TikTok U.S. Data Security (USDS) to strengthen our data protection policies and protocols, further protect our users, and build confidence in our systems, governance, and controls in the U.S. Furthermore, we introduced Project Clover, an initiative to create a specially-designed protective environment for European TikTok user data. With Project Clover, we are designing and implementing a range of new measures to reinforce existing data protections, add new ones, and align our overall approach to data governance with the principle of European data sovereignty.

As we navigate an ever-evolving threat landscape, TikTok remains dedicated to preserving the integrity of our community and strengthening our defenses against adversarial threats. We are committed to upholding the trust placed in us by our users and will continue to invest in robust measures that prioritize user safety, data protection, and regulatory compliance.

Ads

TikTok has strict policies to protect users from fake, fraudulent, or misleading content, including ads. Advertiser accounts and ad content are held to these policies and must follow our Community Guidelines, Advertising Guidelines, and Terms of Service. During the first quarter of 2023, both the total volume of ads removed for violating our advertising policies and the volume of ads removed due to account-level actions decreased. We understand that ensuring the safety of advertisers and users requires ongoing dedication. As such, we are committed to consistently reviewing and further strengthening our systems to swiftly and accurately remove ads that breach our policies. By upholding strict policies, leveraging advanced detection mechanisms, and continuously improving our systems, we strive to foster an advertising experience that is trustworthy, enjoyable, and aligned with the values of our vibrant TikTok community.

Latest data

downloadDownload

Total videos removed/total videos, by quarter

NOTE: Total videos removed represent about 1% of all videos uploaded to TikTok. Video removal volumes referenced in this report include short-form video content (including image-based videos and stories).

Total videos removed/restored, by type and quarter

NOTE: Restored videos are reflected in both the removal volumes and restore volumes in the above chart.

Removal rate, by quarter/policy

NOTE: Proactive removal means identifying and removing a video before it’s reported. Removal within 24 hours means removing the video within 24 hours of it being posted on our platform.

Total video removal, by policy

NOTE: This chart shows the volume of videos removed by policy violation. A video may violate multiple policies and each violation is reflected. In certain rare circumstances, such as emergency situations or hardware outages, we may not always capture the exact video violation category. These videos are not represented in the above chart but are accounted for in absolute numbers throughout this report.

Total video removal and rates, by sub-policy

NOTE: Only videos that have been reviewed and removed by moderators are included in the sub-policy dashboard.

Removal volume and rates, by market

NOTE: This chart shows the fifty markets with the largest volumes of removed videos and represents approximately 90% of overall removal volume.

Human moderation language distribution

NOTE: This chart accounts for the primary language of moderators working on short-form video, livestream, comments, accounts, and TikTok Now. It does not account for moderators who may cover multiple languages. Languages that represent less than 2% of all moderators’ primary languages have been aggregated in the “Other languages” category. Moderation may also take place in additional languages, even if they are not assigned as a primary language among moderators.

Response time to community-reported content

NOTE: Only videos that have been both first reported by users and resulted in a removal action are included. Data reflects the difference between when TikTok received a user report and the time at which the content was removed. Balancing the input and output of human reviews requires matching capacity, training, and content, which occasionally may result in longer review times.

Total account removal, by quarter and reason

NOTE: In addition to removing accounts for our violating Community Guidelines, we remove accounts determined to be spam, along with spam videos posted by those accounts. We also take proactive measures to prevent spam accounts from being created through automated means.

Fake engagement

NOTE: We take action to both remove and prevent likes, followers, and follow requests when we deem the activity to come through automated or inauthentic mechanisms.

Spam account activity

NOTE: When we remove accounts for spam, we also remove videos created by those accounts under our spam policies.

Covert influence operations

Covert influence operations networks identified and removed in Q1 2023Detection sourceAccounts in networkFollowers of network
We assess that this network operated from Russia and targeted primarily Russian audiences. The network was partially created prior to our suspending livestreaming and new content in Russia and extensively used location obfuscation as a tactic. The individuals behind this network used inauthentic identities, including inauthentic news outlets, in order to artificially amplify a pro-Russia viewpoint targeting discourse about the war in Ukraine.Internal1538,326
We assess that this network operated from Russia and targeted Russian audiences. The network extensively used location obfuscation as a tactic to mislead TikTok’s systems. The individuals behind this network used inauthentic accounts to hyper-post content in order to artificially promote Yevgeny Viktorovich Prigozhin’s war film “The Best in Hell”, as well as to amplify pro-Russia viewpoints in the context of the war in Ukraine.Internal25434,110
We assess that this network operated from Poland and targeted Polish audiences. The individuals behind this network created inauthentic identities and hyper-posted comments with similar content in order to artificially promote anti-Russian viewpoints.Internal4140,256
We assess that this network operated from Israel and targeted Israeli audiences. The individuals behind this network used inauthentic identities in order to artificially amplify specific pro-Israeli viewpoints targeting discourse about the ongoing conflict in Palestine in the context of the latest Israeli elections.Internal362168,202
We assess that this network operated from Russia and targeted Moldovan audiences. The individuals behind this network posed as inauthentic news agencies and created inauthentic identities in order to cultivate a follower base on TikTok and then redirect followers off-platform to circumvent the regional TikTok ban in Russia.Internal1155,066
We assess that this network operated from Russia and targeted Russian audiences. The individuals behind this network created inauthentic accounts and hyper-posted content with the aim of artificially amplifying specific pro-Russia viewpoints in the context of the war in Ukraine, abusing the re-share function and using the web application in order to circumvent the regional TikTok ban in Russia.Internal1,351226,838
We assess that this network operated from the UK and Nigeria and targeted Nigerian audiences. The individuals behind this network used inauthentic identities and posed as inauthentic news outlets in order to artificially amplify viewpoints related to the Biafra region, targeting discourse about elections in Nigeria.Internal1547,103
We assess that this network operated from Russia and targeted various European countries such as Germany, Italy, and the UK. The individuals behind this network used impersonation in order to artificially amplify specific viewpoints related to Ukraine’s president Zelensky, the economic sanctions currently imposed on Russia, and Ukrainian refugees.Internal121,480
We assess that this network operated from Ireland and targeted Irish audiences. The individuals behind this network created inauthentic accounts; hyper-posted content with divisive views related to nationalism in Ireland, Japan, Russia, and Taiwan; and hyper-posted comments with similar low-quality content in an attempt to redirect TikTok users off-platform and to intensify social conflict.
Internal7294,743
We assess that this network operated from Malaysia and targeted Malay audiences. The individuals behind this network created inauthentic accounts and hyper-posted identical comments in English and Malay on multiple videos, targeting discourse on elections in Malaysia.Internal175285,511
We assess that this network operated from Ukraine and targeted Ukrainian audiences. The individuals behind this network created inauthentic accounts and hyper-posted content with the aim of artificially amplifying specific viewpoints related to the Ukrainian government and of promoting a positive image of President Zelensky.Internal11990,303
We assess that this network operated from Germany and targeted Egyptian audiences. The individuals behind this network created inauthentic accounts in order to artificially amplify calls to join non-existing protests in Egypt targeting discourse around the current Egyptian government.Internal6368,644

NOTE: Disrupting a covert influence operation requires layered operations, including investigation, removal, and post-mortem analysis. We report the removal of these networks during the quarter in which the full operations process has been completed.

Terminology

  • Networks operated from: Indicates geographic location of network operation based on technical and behavioural evidence from proprietary and open sources; TikTok may not be able to attribute networks to specific entities, individuals, or groups.
  • Detection source: Denoted as internal when presence of activity is identified solely through an internally driven investigation; external detection refers to investigations that originated through an external report which then led to an investigation.
  • Followers of network: Cumulative total number of accounts that followed any account within a network as of the date of that network’s removal.

Ads policy enforcement

NOTE: Ads may be removed either at the individual ad level or in bulk by taking action against an entire advertiser account.

Other reports

Was this helpful?

thumps upYesthumps downNo