Here you can find our previous quarterly transparency reports, dating back to July 2013. If you're not sure what a term means, please see our chapter on Definitions. For more information on how we respond to requests for user information, please see our guidelines for requesting user information.

January–June 2021

At Pinterest, our mission is to bring everyone the inspiration to create a life they love, and it’s our guiding light in drafting our content policies. 

Not everything on the internet is inspiring, so we have guardrails for what’s acceptable on Pinterest and what isn’t allowed. Our moderation practices are always evolving to keep up with new behaviors and trends and to create a more positive corner of the internet for the people on our platform. We continue to invest heavily in measures like machine learning technology to fight policy-violating content on Pinterest and to work with outside experts and organizations to inform our policies and content moderation practices.

We started publishing a biannual transparency report in 2013. Then, starting with Q4 2020, we expanded the report to include new information on the actions we take to uphold our Community guidelines. In this transparency report, you’ll find information on our efforts to moderate user and merchant content, such as the number of policy violations and deactivations. This report also includes insight into the volume of information and deactivation requests we received from law enforcement and government agencies. The data reported here covers the first half of 2021, from January through June 2021.

Pinterest’s industry-leading policies and practices are something we’re proud of. But more importantly, they’re the right thing for the people on our platform. They help to keep Pinterest a more positive and inspiring place online—for example, we prioritize mental health with features like our compassionate search function, and help protect users under our longstanding policy prohibiting medical misinformation. We want to advance the industry on these issues so that—together—we can create a more inspiring internet.

What’s new in this report

This report might look a little different from our last transparency report. For one thing, we know that numbers alone can’t tell the whole story so we’ve added a section on Community safety and wellbeing, which describes some of the ways we’ve invested in our community through policies, products and partnerships this past half. We’re proud of the work we’ve done in this area, and we’re committed to continuing to invest in the health of our community.

We’ve also updated our section on Community guidelines enforcement to be organized by policy category instead of action type. So, for example, in our last report there was a section on Pin deactivations and another for board deactivations—now, there is a section for each policy category that includes the number of Pins and boards deactivated for violating that policy. We made this change to provide a more holistic view of our content moderation actions within a policy category. 

It’s important to note that, even with this new format, we’re still reporting all the same information that we included in our last transparency report. In fact, we’ve also added new information—specifically, data on how we deactivate Pins, whether that’s through automated tools, manual review or a hybrid approach that combines elements of both. Check out the methodology section below for more information on how we calculated this data.

We’re committed to providing greater transparency into how we keep Pinterest safe and inspiring, and we'll continue to iterate on this report going forward.

Community safety and wellbeing

We continue to invest in our policies, products and partnerships to support the safety and wellbeing of our community. Here are just some of the key updates we made in the first half of 2021.

Algorithms and Skin Tone Ranges

As a content distribution platform, we have a responsibility to diminish bias in the algorithms and systems that drive Pinterest. We do this in large part by applying diverse data in our artificial reality (AR) and artificial intelligence (AI) technology. Our focus in these areas can be seen in how we’ve developed features like AR Try on and skin tone ranges that allow users to refine beauty-related searches to find what they’re looking for. Our dataset has diversified over time to help people customize their searches. In the first half of the year, there were millions of searches using skin tone ranges, and in March 2021, we expanded the feature to thirteen additional countries: France, Italy, Spain, Germany, Switzerland, Austria, Netherlands, India, Indonesia, Brazil, Argentina, México, and Colombia. 

Launching the Creator Code

As part of our work to build a positive online space for creators, we introduced the Creator Code, our content commitment designed to educate and empower content creators to fulfill their responsibility to keep Pinterest a positive and inspiring place. The Code outlines core pillars that Creators must accept before posting Idea Pins: Be kind. Check your facts. Do no harm. Watch for triggers. Practice inclusion. 

Simply put, the Code asks Creators to lead with positivity, kindness and inclusion. This collective effort to mitigate negativity is a critical step in our journey to build a positive experience that adds lasting value to the lives of people on Pinterest.

Partnership with National Eating Disorders Association to support body positivity

In February 2021, we supportedNational Eating Disorders Awareness Week in partnership with the National Eating Disorders Association (NEDA), the largest nonprofit organization in the United States dedicated to supporting individuals and families affected by eating disorders.

With this year’s theme, NEDA invited “Every Body to Have a Seat at the Table” by welcoming conversations on awareness, challenging systemic biases and sharing stories from all backgrounds and experiences. We donated ad credits to encourage people to tune into NEDAwareness Week events and the different roundtables taking place. NEDA used a new Pinterest format—Idea Pins—to educate users about eating disorders communication and recovery, and we featured those Idea Pins on the Today Tab to help generate awareness of these topics on our platform.

We continuously work with third party experts and organizations, like NEDA, to get advice and feedback on our policy and enforcement approaches. Collaborations with experts help us properly address policy updates, look for cross-platform trends and combat misinformation.

New initiatives to fight against COVID-19 vaccine misinformation

For several years, we have been at the forefront of combating health misinformation online and ensuring people on Pinterest have access to authoritative information about important health issues, like vaccines. In fact, we were one of the first platforms to implement a policy prohibiting health misinformation in 2017. 

The team introduced our first phase of COVID misinformation efforts in February 2020, including customizing the search experience with results from experts like the World Health Organization (WHO) and Centers for Disease Control (CDC) and prohibiting ads that offered misleading cures or treatments for COVID-19.

In March 2021, we furthered our commitment to combat COVID-19 vaccine misinformation by:

  • Continuing to showcase authoritative content on Expert Search and Today Tab. When people search on Pinterest for topics related to coronavirus, a banner that leads to the WHO COVID-19 resource page appears so users can learn more from trusted sources. Expert Search results also feature relevant content created by doctors and medical organizations like the WHO and the CDC.
  • Partnering with the Association for Healthcare Social Media and Greater Than COVID to create content that educates people on Pinterest about COVID-19 vaccines, generates awareness about the importance of vaccination and dispels myths.
  • Offering paid media, creative resources and support for medical organizations serving underrepresented communities to amplify and elevate reliable third-party COVID-related resources on Pinterest.
  • Providing paid media to the AdCouncil in support of its national COVID-19 vaccine awareness campaign “It’s Up to You.”
  • Community guidelines enforcement

    Pinterest’s Community guidelines are designed to support our mission of inspiration. They govern what we do and don't allow on Pinterest, and all users must abide by them. 

    We have additional guidelines for merchants and advertisers to set clear expectations about what is and is not acceptable for product Pins and advertisements. These guidelines include especially high standards that are for the safety of all audiences who use Pinterest. We believe you can't feel inspired if you don't first feel safe.

    To help us cultivate a safe and inspired community, we develop and enforce content policies that help ensure our platform is a positive place where people can find real-life ideas for what to try next, cook next, wear next, or do next. We work hard to identify and deactivate harmful content from our site, and our content policies and moderation practices are always evolving to keep up with new behaviors and trends and to create a more positive corner of the internet for users. 

    We may block, limit the distribution of or deactivate content and the accounts, individuals and groups that create or spread that content, based on how much harm it poses. In the event that a user believes a deactivation was in error, Pinterest provides options to appeal the deactivation.

    MethodologyDistinct images and Pins deactivated

    Every day, millions of people all over the world come to Pinterest to create, discover and save new ideas that are shared in Pins. To understand how we approach content moderation, it’s helpful to differentiate between two types of Pins: organic Pins and ads. Our Community guidelines apply to both.

    Organic Pins include all Pins created and saved on Pinterest that are not promoted as ads. For example, this could include merchants’ product Pins, which aren’t always ads, and may appear organically to people who are searching for products on Pinterest. We have additional requirements, like that the Pin image and description must accurately represent the product, for merchants and their product Pins. All types of organic Pins are included in our transparency reports.

    Ads are Pins that businesses pay to promote. We have additional policies for advertisers that hold ads and advertisers to even higher standards. Ad policies are enforced differently than organic content, and are not included in these transparency reports.

    Much of the content on Pinterest has been saved repeatedly, meaning that the same image may appear in multiple Pins. So when it comes to reporting content moderation for organic Pins, we include the number of Pins deactivated as well as the number of distinct images deactivated to provide greater insight into our moderation practices for this type of content. 

    We report boards and accounts deactivated separately. To avoid double-counting deactivations, our count of distinct images and Pins deactivated does not include those on boards or from user accounts that were deactivated.

    How we deactivate Pins

    We deactivate policy-violating Pins through automated tools, manual review and a hybrid approach that combines elements of both. 

    Automated deactivations. Our automated tools use a combination of signals to identify and take action against potentially violating content. Our machine learning models assign scores to each image added to our platform. Using those scores, our automated tools can then apply the same enforcement decision to other Pins containing the same images.

    Manual deactivations. We manually deactivate Pins through our human review process. Pins deactivated through this process may include those identified internally and those reported to us by third parties. It also includes the Pins that are reviewed and deactivated by one of our team members after a user report.

    Hybrid deactivations. Hybrid deactivations include those where a human determines that a Pin violates policy, and automated systems help expand that decision to enforce against machine-identified matching Pins. Depending on the prevalence of matching Pins, a hybrid deactivation may result in a number of Pins deactivated or none at all. 

    The mechanisms used to address different potential policy violations may vary based on the state of available technology, the volume of violative content and other factors such as the complexity of evaluation. We continue to iterate and evolve our tools and expect ongoing improvements going forward.

    Reach of Pins deactivated for violating policy

    People often ask, how many users saw a Pin before it was deactivated? In most cases, the answer is: not a lot. 

    For example, 70% of Pins that we deactivated for civic misinformation in Q2 were never seen by users in that reporting period—even with more than 400 million people visiting Pinterest per month.

    Reach is one of our key indicators of user experience. To calculate this metric, we start by looking at each policy-violating Pin deactivated in a reporting period. Then we count the number of unique users that saw each of those Pins during the reporting period for at least 1 second, before it was deactivated. Reach for a policy category may not add up to 100% due to rounding.

    Actioned user reports

    Users can report Pins they find objectionable by clicking on the three small dots on any Pin and hitting “Report Pin.” Once we confirm a policy violation and take action, we consider the report an actioned user report.

    The total number of actioned user reports—in other words, reports found to have correctly identified policy violations—is another key indicator of user experience on Pinterest. So is the number of reporters: fewer than 0.01% of monthly active users reported a Pin that resulted in a Pin deactivation in Q2.

    Boards deactivated

    When users find Pins they like or want to come back to, they save them to boards that they’ve created. Over time, our users have created billions of boards. 

    When a board is deactivated for violating policy, all the Pins on that board are also deactivated. Similarly, when we deactivate an entire account, that user’s boards are also deactivated. To avoid double-counting deactivations, our count of boards deactivated does not include those from user accounts that were deactivated.

    Accounts deactivated

    Think of “accounts” on Pinterest as “profiles” or, most often, individual users. If someone saves Pins and creates boards, that content becomes associated with their account. This is also true for business, advertiser, merchant and creator accounts. 

    Any account may be deactivated for violating our policies. When an account is deactivated, all of their Pins and boards are also deactivated. That means that if you search for them or click on an old link to their profile, that profile won’t show up anymore. Their Pins won’t appear anywhere on Pinterest. And the deactivated user won’t be able to access their own Pins or boards, either.

    Account appeals and reinstatements

    If people believe their accounts have been deactivated by mistake, they can follow an appeals process to have their accounts reinstated. We review appeal requests and grant the appeal if we determine we made a mistake, or in some cases to give people another chance to abide by our Community guidelines.

    We also process appeals for deactivated Pins and boards and expect to include that data in reports in the future.

    Reporting periods

    Our reporting on Community guidelines enforcement covers the first half of 2021, from January to June 2021. We’ve split that time into two reporting periods: Q1 and Q2. 

    Q1 covers the first quarter of the year, from January through March 2021, while Q2 covers the second quarter of the year, from April through June 2021. Sometimes we also refer to Q4 2020 (October to December 2020) when we’re talking about quarter-over-quarter trends.

    Adult content

    Pinterest isn’t a place for pornography. We deactivate content that violates our adult content policy and do not deliver results for searches determined to violate this policy. We have established guidelines to help differentiate between pornography and other mature content such as content related to sexual health, breastfeeding, mastectomies, art and education.

    Recent trends

    In Q1 2021, we launched new machine learning tools to detect and deactivate boards for violating our adult content policy. As a result, we saw an increase in the number of boards and accounts deactivated. Looking at Q2 2021 compared to Q4 2020, there was also a 15% decrease in actioned user reports per monthly active user—one of our key indicators of user experience—that we attribute to this effort.

    Our content policies and moderation practices are always evolving to better keep our platform safe for all users. So far, we’ve focused our use of these new machine learning tools on enforcing against our adult content policy, and we’re currently exploring how they may work for other types of policy violations. 

    Content enforcementReach* of Pins deactivated for adult content
      Q1 2021 Q2 2021
    Seen by 0 people 80% 81%
    Seen by 1-10 people 15% 15%
    Seen by 10-100 people 3% 3%
    Seen by 100+ people 1.6% 1.3%

    * To calculate reach, we start by looking at each policy-violating Pin deactivated in a reporting period. Then we count the number of unique users that saw each of those Pins during the reporting period for at least 1 second, before it was deactivated. Quarterly percentages may not add up to 100% due to rounding.

    In Q1 2021, we deactivated 2,029,934 distinct images, which comprised 47,120,204 Pins, for violating this policy. Of these Pins, 80% were never seen by users in this reporting period. We deactivated fewer than 1% manually and more than 99% with hybrid tools. We actioned 128,015 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 179,994 boards and 17,659 accounts. We received 2,335 account appeals and reinstated 1,634 accounts.

    In Q2 2021, we deactivated 1,486,098 distinct images, which comprised 46,130,733 Pins, for violating this policy. Of these Pins, 81% were never seen by users in this reporting period. We deactivated fewer than 1% manually and more than 99% with hybrid tools. We actioned 136,218 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 131,664 boards and 13,468 accounts for violating this policy. We received 2,398 account appeals and reinstated 1,442 accounts.

    Adult sexual services

    We prohibit adult sexual services on Pinterest, and the amount of such content we find on the platform is low. Adult sexual services may involve commercial sex work and sexual or physical exploitation or trafficking, such as escort services, prostitution, sex cams, or sex- or nudity-for-money services.

    Recent trends

    Sometimes Pins do not have content that would violate our adult content policy but may still contain offers for adult sexual services. When we identify Pins like that, we deactivate them for violating our adult sexual services policy. 

    In Q2 2021, we deactivated 1,263 Pins for violating this policy, and of those, 78% were seen by fewer than 100 users in that reporting period.

    Content enforcementReach of Pins deactivated for adult sexual services
      Q1 2021 Q2 2021
    Seen by 0 people 1.5% 0.7%
    Seen by 1-10 people 41% 45%
    Seen by 10-100 people 28% 32%
    Seen by 100+ people 29% 22%

     

    In Q1 2021, we deactivated 520 distinct images, which comprised 533 Pins, for violating this policy. Of these Pins, 71% were seen by fewer than 100 users in this reporting period. We deactivated 100% of Pins manually. We actioned 536 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 348 boards and 271 accounts. We received 11 account appeals and reinstated 3 accounts.

    In Q2 2021, we deactivated 1,245 distinct images, which comprised 1,263 Pins, for violating this policy. Of these Pins, 78% were seen by fewer than 100 users in this reporting period. We deactivated 100% of Pins manually. We actioned 613 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 708 boards and 500 accounts for violating this policy. We received 14 account appeals and reinstated 4 accounts.

    Civic misinformation

    We introduced our civic misinformation policy in 2020 as an addition to several existing Community guidelines on misinformation and disinformation. This policy prohibits false or misleading content on Pinterest that impedes an election’s integrity or an individual's or group's civic participation, including registering to vote, voting and being counted in a census.

    Recent trends

    As part of our efforts to maintain a safe platform during the US presidential transition, there was a relative increase in the number of boards and accounts deactivated for violating this policy in Q1 2021.

    Content enforcementReach of Pins deactivated for civic misinformation
      Q1 2021 Q2 2021
    Seen by 0 people 80% 70%
    Seen by 1-10 people 16% 26%
    Seen by 10-100 people 3% 3%
    Seen by 100+ people 0.9% 0.8%

     

     

    In Q1 2021, we deactivated 2,044 distinct images, which comprised 6,895 Pins, for violating this policy. Of these Pins, 80% were never seen by users in this reporting period. We deactivated 96% of Pins manually and 4% with hybrid tools. We actioned 111 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 546 boards and 45 accounts. We received 13 account appeals and reinstated 11 accounts.

    In Q2 2021, we deactivated 349 distinct images, which comprised 2,316 Pins, for violating this policy. Of these Pins, 70% were never seen by users in this reporting period. We deactivated 16% of Pins manually and 84% with hybrid tools. We actioned 118 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 92 boards and 9 accounts for violating this policy. We received 4 account appeals and reinstated 3 accounts.

    Conspiracy theories

    Our conspiracy theory policy is one of several policies in our Community guidelines that prohibit misinformation and disinformation on Pinterest. Conspiracy theories include content that turns or encourages turning individuals, groups of people, places or organizations into targets of harassment or physical violence, such as hate-based conspiracy theories and misinformation about mass atrocities.

    Recent trends

    Sometimes, we perform a sweeping cleanup across the platform that can cause a temporary spike in take-downs of certain types of content. The ultimate outcome is of course a good thing: a cleaner, safer, more inspiring place for everyone. 

    We saw this with conspiracy theories in Q2 when we performed a cleanup across the platform, as part of our usual content enforcement efforts. This generated a temporary spike in Pins deactivated. This rise is not due to any known increase of violative content, and more importantly 95% of those Pins were never seen by users during that reporting period.

    Content enforcementReach of Pins deactivated for conspiracy theories
      Q1 2021 Q2 2021
    Seen by 0 people 92% 95%
    Seen by 1-10 people 6% 5%
    Seen by 10-100 people 1% 0.3%
    Seen by 100+ people 0.3% 0.1%

     

    In Q1 2021, we deactivated 24,134 distinct images, which comprised 166,189 Pins, for violating this policy. Of these Pins, 92% were never seen by users in this reporting period. We deactivated 2% of Pins manually and 98% with hybrid tools. We actioned 217 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 592 boards and 116 accounts. We received 20 account appeals and reinstated 11 accounts.

    In Q2 2021, we deactivated 16,204 distinct images, which comprised 1,148,947 Pins, for violating this policy. Of these Pins, 95% were never seen by users in this reporting period. We deactivated fewer than 1% manually and more than 99% with hybrid tools. We actioned 318 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 451 boards and 81 accounts for violating this policy. We received 13 account appeals and reinstated 5 accounts.

    Dangerous goods and activities

    Pinterest isn’t a place for trading, selling, or engaging in dangerous goods or activities. Dangerous goods are products or substances that might cause harm when used, altered or manufactured irresponsibly. Dangerous activities—including displaying or encouraging dangerous activities—are also not welcome on Pinterest.

    Recent trends

    We performed a sweeping cleanup across the platform for dangerous goods and activities content in Q1 and Q2 2021, leading to a temporary spike in take-downs of this type of content. The ultimate outcome of work like this is a cleaner, safer, more inspiring place for everyone. 

    This effort, applied to organic Pins and not promoted ads or products, generated a temporary spike in Pins deactivated for violating this policy. This rise is not due to any known sustained increase of violative content.

    A smaller portion of the increase in Pins deactivated in Q1 2021 was driven by the detection and deactivation of Pins related to a harmful weight-control drug, which also led to the deactivation of a significant number of boards and accounts. 

    Content enforcementReach of Pins deactivated for dangerous goods and activities
      Q1 2021 Q2 2021
    Seen by 0 people 95% 97%
    Seen by 1-10 people 4% 2%
    Seen by 10-100 people 0.6% 0.2%
    Seen by 100+ people 0.4% 0.1%

     

    In Q1 2021, we deactivated 108,534 distinct images, which comprised 839,389 Pins, for violating this policy. Of these Pins, 95% were never seen by users in this reporting period. We deactivated 5% manually, fewer than 1% automatically and 95% with hybrid tools. We actioned 3,136 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 2,344 boards and 403 accounts. We received 14 account appeals and reinstated 4 accounts.

    In Q2 2021, we deactivated 938,363 distinct images, which comprised 12,286,669 Pins, for violating this policy. Of these Pins, 97% were never seen by users in this reporting period. We deactivated fewer than 1% of Pins manually, fewer than 1% automatically and more than 99% with hybrid tools. We actioned 3,394 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 1,814 boards and 501 accounts for violating this policy. We received 17 account appeals and reinstated 5 accounts.

    Graphic violence and threats

    We prohibit graphic violence and threats on Pinterest. This includes most content that shows the use of violence, threats and language that glorifies violence.

    Recent trends

    There was a relative increase in the amount of content deactivated for violating our graphic violence and threats policy during the US presidential transition in Q1 2021.

    Content enforcementReach of Pins deactivated for graphic violence and threats
      Q1 2021 Q2 2021
    Seen by 0 people 96% 26%
    Seen by 1-10 people 3% 28%
    Seen by 10-100 people 0.4% 21%
    Seen by 100+ people 0.5% 26%

     

    In Q1 2021, we deactivated 3,376 distinct images, which comprised 254,455 Pins, for violating this policy. Of these Pins, 96% were never seen by users in this reporting period. We deactivated 2% of Pins manually and 98% with hybrid tools. We actioned 873 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 1,184 boards and 243 accounts. We received 15 account appeals and reinstated 10 accounts.

    In Q2 2021, we deactivated 4,005 distinct images, which comprised 7,173 Pins, for violating this policy. Of these Pins, 74% were seen by fewer than 100 users in this reporting period. We deactivated 79% of Pins manually and 21% with hybrid tools. We actioned 1,848 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 767 boards and 72 accounts for violating this policy. We received 29 account appeals and reinstated 13 accounts.

    Harassment and criticism

    Pinterest is not a place for harassment and criticism, and we prohibit content that insults, hurts or antagonizes individuals or groups of people. This includes manipulated images intended to degrade or shame, shaming people for their bodies or assumed sexual or romantic history, sexual remarks about people’s bodies, solicitations or offers of sexual acts, and mocking someone for experiencing sadness, grief, loss or outrage.

    Recent trends

    We try to be very mindful about when we use hybrid tools and—more importantly—how to scale them. Remember, hybrid deactivations include those where a human determines that a piece of content violates policy, and automated systems expand that decision to enforce against machine-identified matching Pins. For example, a hybrid deactivation may be appropriate if we determine that the image in a Pin violates policy, but it may not be appropriate if the same image might be allowed in other situations, or if the initial Pin is deactivated for reasons other than the image it shows.

    In Q2 2021, the vast majority of Pins deactivated for violating our harassment and criticism policy were the result of a hybrid deactivation that included a small handful of misclassified images. We reactivated those deactivated Pins once we spotted the error and have made improvements to the process to minimize the chances that it will happen again.

    We’ve included those false positives in the Q2 enforcement data, but we excluded them from the reach metric for this policy in an effort to provide more accurate insight into the number of users who saw a Pin that actually violates this policy before the Pin was deactivated. For Q2, 79% of such Pins were never seen by people in that reporting period.

    Content enforcementReach of Pins deactivated for harassment and criticism
      Q1 2021 Q2 2021
    Seen by 0 people 87% 79%
    Seen by 1-10 people 10% 18%
    Seen by 10-100 people 2% 3%
    Seen by 100+ people 1% 1%

     

    * See the Recent trends section above for details on how we calculated reach for this policy in Q2 2021.

    In Q1 2021, we deactivated 5,540 distinct images, which comprised 124,713 Pins, for violating this policy. Of these Pins, 87% were never seen by users in this reporting period. We deactivated 11% of Pins manually and 89% with hybrid tools. We actioned 4,483 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 977 boards and 594 accounts. We received 57 account appeals and reinstated 43 accounts.

    In Q2 2021, we deactivated 7,238 distinct images, which comprised 1,238,782 Pins, for violating this policy. We determined that a small handful of these distinct images, and their more than 990,000 machine-identified matching Pins, were incorrectly deactivated, and we reinstated that content after spotting the error. Of the Pins that we believe were correctly deactivated, 79% were never seen by users in this reporting period. We deactivated fewer than 1% of Pins manually and more than 99% with hybrid tools. We actioned 5,383 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 1,025 boards and 292 accounts for violating this policy. We received 33 account appeals and reinstated 22 accounts.

    Hateful activities

    We prohibit hateful content and the people and groups that promote hateful activities on Pinterest. Hateful activities include slurs and negative stereotypes, caricatures and generalizations, as well as support for hate groups and people promoting hateful activities.

    Recent trends

    We believe all people deserve a safe space to cultivate their interests and seek inspiration—regardless of their actual or perceived race, color, caste, ethnicity, immigration status, national origin, religion or faith, sex or gender identity, sexual orientation, disability or medical condition.

    In Q1 and Q2 2021, we saw a relative decrease in boards and accounts deactivated compared to Q4 2020. We deactivated fewer Pins in Q2, and 74% of those Pins were seen by fewer than 100 users in that reporting period.

    Content enforcementReach of Pins deactivated for hateful activities
      Q1 2021 Q2 2021
    Seen by 0 people 70% 22%
    Seen by 1-10 people 13% 34%
    Seen by 10-100 people 7% 18%
    Seen by 100+ people 11% 26%

     

    In Q1 2021, we deactivated 2,487 distinct images, which comprised 8,823 Pins, for violating this policy. Of these Pins, 70% were never seen by users in this reporting period. We deactivated 30% of Pins manually and 70% with hybrid tools. We actioned 862 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 665 boards and 82 accounts. We received 22 account appeals and reinstated 13 accounts.

    In Q2 2021, we deactivated 3,418 distinct images, which comprised 6,086 Pins, for violating this policy. Of these Pins, 74% were seen by fewer than 100 users in this reporting period. We deactivated 60% of Pins manually and 40% with hybrid tools. We actioned 1,948 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 739 boards and 126 accounts for violating this policy. We received 23 account appeals and reinstated 12 accounts. 

    Medical misinformation

    Our medical misinformation policy is one of several policies in our Community guidelines that prohibit misinformation and disinformation on Pinterest. This policy  prohibits medically unsupported health claims that risk a user’s health or wider public health and safety, including the promotion of false cures, anti-vaccination advice, or misinformation about public health or safety emergencies. We rely on information from nationally and internationally recognized institutions, including the CDC and WHO, to help us determine if content violates these guidelines.

    Recent trends

    Pinterest is deeply committed to combating health misinformation. As COVID-19 vaccines were made more widely available around the world, we deactivated content that violated our medical misinformation policy and worked closely with health experts to ensure we stayed up-to-date on the latest developments. We saw a relative increase in the number of Pins deactivated in Q2 2021, and 92% of these Pins were seen by fewer than 100 users in this reporting period.

    Content enforcementReach of Pins deactivated for medical misinformation
      Q1 2021 Q2 2021
    Seen by 0 people 74% 65%
    Seen by 1-10 people 16% 19%
    Seen by 10-100 people 3% 8%
    Seen by 100+ people 8% 8%

     

    In Q1 2021, we deactivated 4,256 distinct images, which comprised 11,097 Pins, for violating this policy. Of these Pins, 74% were never seen by users in this reporting period. We deactivated 48% of Pins automatically, 12% manually and 40% with hybrid tools. We also deactivated 473 boards and 35 accounts. We received 7 account appeals and reinstated 5 accounts.

    In Q2 2021, we deactivated 4,869 distinct images, which comprised 19,852 Pins, for violating this policy. Of these Pins, 65% were never seen by users in this reporting period. We deactivated 31% of Pins automatically, 30% manually and 39% with hybrid tools. We also deactivated 462 boards and 34 accounts for violating this policy. We received 11 account appeals and reinstated 6 accounts.

    Self-injury and harmful behavior

    Combating self-harm is a priority for us as we strive to ensure Pinterest plays a positive role in people’s lives. Content that displays, rationalizes or encourages suicide, self-injury, eating disorders or substance abuse isn’t welcome on our platform. This includes self-harm instructions, sensitive imagery, and suicidal thinking and quotes. 

    Our ongoing partnership with Samaritans, an organization dedicated to reducing feelings of isolation and disconnection that can lead to suicide, continues to help the industry tackle self-harm content and make the internet a safer place for everyone. In 2019, we deepened our relationship with the Samaritans by joining their newly formed industry-wide program. Partnerships like this help us support expert research and development within the field of mental health so that more people in need can find support, no matter what app or website they use. 

    Recent trends

    In Q1 and Q2, we continued investing in work to improve content moderation for self-harm content and provided support millions of times to those in need through our compassionate search experience.

    In Q1 2021, hybrid deactivations resulted in more Pins deactivated, but fewer distinct images deactivated, than they did in Q2. More importantly, 99% of the Pins deactivated in Q1 were seen by fewer than 100 people in that reporting period. 

    Content enforcementReach of Pins deactivated for self-injury and harmful behavior
      Q1 2021 Q2 2021
    Seen by 0 people 86% 88%
    Seen by 1-10 people 12% 8%
    Seen by 10-100 people 2% 2%
    Seen by 100+ people 1% 2%

     

    In Q1 2021, we deactivated 2,945 distinct images, which comprised 302,918 Pins, for violating this policy. Of these Pins, 86% were never seen by users in this reporting period. We deactivated 1% of Pins manually and 99% with hybrid tools. We actioned 1,066 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 545 boards and 79 accounts. We received 13 account appeals and reinstated 12 accounts.

    In Q2 2021, we deactivated 3,824 distinct images, which comprised 81,444 Pins, for violating this policy. Of these Pins, 88% were never seen by users in this reporting period. We deactivated 10% of Pins manually and 90% with hybrid tools. We actioned 1,567 user reports that resulted in a Pin deactivated for violating this policy. We also deactivated 828 boards and 123 accounts for violating this policy. We received 28 account appeals and reinstated 25 accounts.

    Spam

    We want the inspiration and ideas on Pinterest to be high-quality and useful, so we deactivate spam when we find it. The goal of spammers is to make money, and the best way to do this is to spam at scale. It’s a numbers game: one million spam emails are much more effective than one spam email. The same kinds of spamming efforts can happen on any content distribution platform, including Pinterest. Platforms’ interactions with spammers are generally both adversarial and iterative with users who intentionally try to evade the system and continually update their techniques. 

    Recent trends

    At Pinterest, we use the latest in machine learning technology to build automated models that swiftly detect and act against spam of all kinds. We not only use the latest modeling techniques but also iterate on these models at regular intervals by adding new data and exploring new technical breakthroughs to either maintain or improve their performance over time to effectively address spam. Given the adversarial, iterative nature of fighting spam, content enforcement numbers may change quarter-to-quarter, especially after a large attack.

    Content enforcementReach of Pins deactivated for spam
      Q1 2021 Q2 2021
    Seen by 0 people 65% 61%
    Seen by 1-10 people 15% 20%
    Seen by 10-100 people 13% 12%
    Seen by 100+ people 7% 7%

     

    In Q1 2021, we deactivated 336,821 distinct images, which comprised 1,042,960 Pins, for violating this policy. Of these Pins, 65% were never seen by users in this reporting period. We deactivated less than 1% of Pins manually and more than 99% automatically. We also deactivated 1 board and 4,420,965 accounts. We received 112,139 account appeals and reinstated 84,541 accounts.

    In Q2 2021, we deactivated 118,054 distinct images, which comprised 378,490 Pins, for violating this policy. Of these Pins, 61% were never seen by users in this reporting period. We deactivated less than 1% of Pins manually and more than 99% automatically. We also deactivated 3,104,408 accounts for violating this policy. We received 80,624 account appeals and reinstated 61,050 accounts.

    Law enforcement and government agency requests

    This section provides insight into the volume of information and deactivation requests received from law enforcement and government agencies. 

    Please note that this data encompasses requests from January through June 2021. For more information on how we respond to requests for account information, refer to our Law enforcement guidelines.

    Government information requests

    Pinterest receives legal requests from law enforcement and government agencies for Pinterest account information. We diligently review each request, and only produce data for those that meet the requirements of law and our policies. Our policy is to notify users of government requests for their information prior to disclosing, unless we are prohibited by law or in exceptional circumstances.

    United States
    Types Requests Some Information Produced Accounts Identified Accounts Notified**
    Subpoena 58 45 113 0
    Court Order 2 2 2 1
    Warrant 21 17 26 1
    Other* 5 1 3 0
    Total 86 65 144 2

    * Law enforcement requests such as wiretap orders, pen registers, trap and trace, and emergency disclosure requests.
    ** The account owner was notified before production.

    International
    Country Types Requests Some Information Produced Accounts Identified Accounts Notified**
    Australia Other* 1 0 2 0
    Costa Rica Other* 2 0 0 0
    Germany Other* 5 0 3 0
    India Other* 2 0 2 0
    Ireland Other* 1 0 0 0
    Ireland Court Order 1 1 1 1
    Israel Other* 1 1 1 0
    South Korea Other* 2 0 8 0
    Mexico Other* 1 0 1 0
    Switzerland Other* 1 0 1 0
    United Kingdom Other* 1 0 1 0
    Total   18 2 20 1

    * Law enforcement requests such as wiretap orders, pen registers, trap and trace, and emergency disclosure requests.
    ** The account owner was notified before production.

    National security requests*
    Time period No. of requests
    January through June 2021 0-249

    * Any national security letters and orders issued under the US Foreign Intelligence Surveillance Act for user information.

    Government content deactivation requests

    We sometimes receive requests from government agencies to deactivate content on Pinterest that may be illegal in their country and/or a violation of our Community guidelines

    We review the requests to determine if the content identified violates our Community guidelines. Our teams take action on violations, ranging from deactivating the content globally to restricting access to the content within the relevant country if it violates local law but does not violate our policies. 

    Pinterest received a total of 7,884 requests from January to June 2021. We deactivated content for 5,396 of those requests for violating our Community guidelines and restricted content on 1,034 requests. Content for the remaining 1,454 requests was inactive by the time it was reviewed in response to the government removal request. This can happen when, for instance, the content was deactivated in response to a user report.

    Country Requests Community Guidelines Deactivations* Local Law Deactivation** Inactive***
    France 1 1 0 0
    Germany 18 12 6 0
    India 36 2 34 0
    Japan 2 2 0 0
    South Korea 5,278 4,291 885 102
    Portugal 9 9 0 0
    Russia 2,402 1,041 63 1,298
    Turkey 138 38 46 54
    Total 7,884 5,396 1,034 1,454

    * Content violated our Community guidelines and was removed from the platform.
    ** Content was reported by a government agency but did not violate our Community guidelines and was restricted from appearing only in the country where the request originated, based on local law.
    *** Content was no longer available on the platform by the time it was reviewed in response to the government removal request.

    Child safety

    Pinterest does not tolerate child sexual abuse material (CSAM). We have a strict no-tolerance policy for any content—imagery or text—that exploits or endangers minors. Detecting and deactivating this content is extremely important to us and we work closely with the National Center for Missing and Exploited Children (NCMEC) to combat this type of activity.

    Pinterest proactively identifies CSAM images and videos using matching tools including PhotoDNA and machine learning tools to detect additional possible CSAM. Our team of specialists is trained to identify and review CSAM, and was responsible for 890 CyberTipline reports to NCMEC from January to June 2021.

    Time period Cybertip reports
    January through June 2021 890
    Building toward a safer internet

    Creating the most positive space online doesn't happen by accident: It happens through proactive policy and product decisions. We have industry-leading positions on content safety that are informed by inputs and advice from outside experts, civil society and government. We also invest heavily in measures like machine learning technology to maintain a safe and positive space for the people on Pinterest. We’re proud of what we’re doing to keep Pinterest safe and to move the broader industry forward. 

    Let’s create a safer, more inspired internet, together.

    July–December 2020

    At Pinterest, our mission is to bring everyone the inspiration to create a life they love, and it’s our guiding light in drafting our content policies. Not everything on the internet is inspiring, so we have guardrails for what’s acceptable on Pinterest and what isn’t allowed. 

    We’re committed to providing greater transparency into how we keep Pinterest safe and inspiring. In this transparency report, you’ll get insight into how many information and deactivation requests we received from law enforcement and government agencies between July and December 2020. We’ve reported on this information twice per year since 2013.

    But this year, we’re expanding the report to share more information. Now, our bi-annual transparency report will also include data on the actions we take to moderate user and merchant content on Pinterest beyond those requested by law enforcement, such as the number of policy violations and deactivations. This release specifically covers the actions we took under our policies between October and December 2020. We'll continue to iterate on this report going forward. 

    Pinterest’s industry-leading policies and practices are something we’re proud of. But more importantly, they’re the right thing for the people on our platform. They help to keep Pinterest a more positive and inspiring place online—for example, our longstanding policy prohibiting medical misinformation, which also challenged the industry to go further. We want to advance the industry on these issues so that—together—we can create a more inspiring internet.

    Our content moderation philosophy

    Pinterest’s Community guidelines are designed to keep our platform safe for all users. They govern what we do and don't allow on Pinterest, and all users must abide by them. 

    We have additional guidelines for merchants and advertisers to set clear expectations about what is and is not acceptable for product Pins and advertisements. We have especially high standards for safety for all audiences who use Pinterest: consumers, advertisers, creators, merchants and more. We believe you can't feel inspired if you don't first feel safe.

    To help us cultivate a safe and inspired community, we develop and enforce content policies to ensure that our platform is a positive place for people to find inspiration for their next new recipe or home renovation. We work hard to identify and deactivate harmful content from our site, and our content policies and moderation practices are always evolving. 

    We may block, limit the distribution of or deactivate content and the accounts, individuals and groups that create or spread that content, based on how much harm it poses. In the event that a user believes a deactivation was in error, Pinterest provides options to appeal the deactivation.

    Pins

    Every day, millions of people all over the world come to Pinterest to create, discover and save new ideas that are shared in Pins. To understand how we approach content moderation, it’s helpful to differentiate between two types of Pins: organic Pins and ads. Our Community guidelines apply to both.

    Organic Pins include all Pins created and saved on Pinterest that are not promoted as ads. For example, this could include merchants’ product Pins, which aren’t always ads, and may appear organically to people who are searching for products on Pinterest. We have additional requirements, like disclosing shipping and return policies, for merchants and their product Pins. All types of organic Pins are included in this report.

    Ads are Pins that advertisers have chosen to promote as ads on Pinterest. We have additional policies for advertisers that set higher standards for the quality of ads. Because ad content is enforced differently than organic content, it is not included in this report.

    Reach of policy-violating Pins

    People often ask: before a Pin is deactivated for violating policy, how many people saw it? In most cases, the answer is: not a lot. 

    For an example, let’s look at Pins deactivated for medical misinformation during this reporting period. 85% of Pins that we deactivated for medical misinformation were actually never seen by users in this reporting period—even with more than 440 million people visiting Pinterest per month.

    Reach of deactivated Pins, Oct–Dec 2020*

    Policy Seen by  0 people Seen by <10 people Seen by 10-100 people Seen by 100+ people
    Adult content 76% 17% 4% 2%
    Adult sexual services 10% 25% 28% 36%
    Civic misinformation 64% 27% 5% 3%
    Conspiracy theories 91% 7% 0.9% 0.3%
    Dangerous goods and activities 50% 37% 6% 5%
    Graphic violence and threats 50% 13% 14% 22%
    Harassment and criticism 78% 16% 3% 3%
    Hateful activities 73% 10% 6% 10%
    Medical misinformation 85% 7% 1% 6%
    Self-injury and harmful behavior 90% 8% 0.8% 0.5%
    Spam 85% 5% 3% 6%

    * Calculated based on the number of unique users that saw a policy-violating Pin between October and December 2020 for at least 1 second, before it was deactivated. Rows may not add up to 100% due to rounding.

    Actioned user reports

    Users can report any content they find objectionable by clicking on the three small dots on any Pin and hitting “Report Pin.” Once we confirm it’s a policy violation and take action on the reported content, we consider the report an actioned user report.

    The total number of actioned user reports tells us a lot about the user experience on Pinterest. So does the number of reporters: less than 0.02% of monthly active users reported a Pin that resulted in a Pin deactivation in this reporting period. 

    While the majority of actioned user reports concerned adult content, the number of users that reported that content represent a very small fraction of the people on Pinterest. More importantly, not a lot of people saw the adult content, regardless of whether it was deactivated due to a user report or another enforcement mechanism. In fact, 98% of adult content that was deactivated on Pinterest was seen by fewer than 100 people during the reporting period.

    Actioned user reports that resulted in a deactivated Pin, Oct–Dec 2020

    Policy* Actioned Reports
    Adult content 155,313
    Adult sexual services 749
    Civic misinformation 782
    Conspiracy theories 786
    Dangerous goods and activities 2,296
    Graphic violence and threats 825
    Harassment and criticism 3,865
    Hateful activities 421
    Self-injury and harmful behavior 1,854

    * For some policies, actioned user reports are not a useful metric, for instance because the primary enforcement mechanism involves processing reports in aggregate. For these policies, the reach of deactivated Pins is a more relevant indicator of the user experience.

    Pin deactivations

    There are more than 300 billion Pins on Pinterest, and each of those Pins also has an image associated with it. Just because two Pins show the same image doesn't mean that we count them as the same Pin within our systems, or even that the image came from the same source. 

    This is important when it comes to content moderation: if we determine that the image in one Pin is policy-violating, our tools need to be able to detect and act on matching images amongst the billions of other Pins on Pinterest. So while we detect and deactivate a lot of Pins, those Pins comprise a much smaller number of distinct images. That’s why we’re sharing the number of deactivations for both distinct images and Pins. Each provides a different kind of insight into our moderation practices for this type of content. 

    For example, we took proactive steps to moderate both new and pre-existing content leading up to the US election in November 2020. Those efforts included deactivating a lot of distinct images that violated our conspiracy theory policy. In doing so, we also deactivated the Pins that used those images.

    Distinct image and Pin deactivations, Oct–Dec 2020*

    Policy Distinct images Total Pins
    Adult content 2,100,253 49,855,681
    Adult sexual services 707 714
    Civic misinformation 3,238 15,809
    Conspiracy theories 52,863 1,512,221
    Dangerous goods and activities 5,501 42,310
    Graphic violence and threats 1,754 3,750
    Harassment and criticism 3,763 46,371
    Hateful activities 1,980 8,397
    Medical misinformation 5,938 18,184
    Self-injury and harmful behavior 3,499 175,584
    Spam 1,378,472 3,375,169

    * Does not include distinct images or Pins that were deactivated because they were on a board that was deactivated, or belonged to a user that was deactivated.

    Boards

    When people find Pins they like or want to come back to they save them to boards that they’ve created. Over time, people have created more than 6 billion boards. 

    Pinterest deactivates boards if a predetermined amount of content on that board has been identified as policy-violating. We’re being intentional about not sharing specific numbers for this, because that information can be used to undermine our efforts at content moderation. The most important takeaways about board moderation are that Pinterest will deactivate boards that are determined to be policy-violating, and that when a board is deactivated, all the Pins on that board are also deactivated. 

    Board deactivations, Oct–Dec 2020*

    Policy Boards
    Adult content 50,767
    Adult sexual services 488
    Civic misinformation 456
    Conspiracy theories 1,877
    Dangerous goods and activities 956
    Graphic violence and threats 381
    Harassment and criticism 795
    Hateful activities 4,604
    Medical misinformation 345
    Self-injury and harmful behavior 898
    Spam 2

    * Does not include boards that were deactivated because they belonged to a user account that was deactivated.

    Accounts

    Think of “accounts” on Pinterest as “profiles” or, most often, individual users. If someone saves Pins and creates boards, that content becomes associated with their account. This is also true for business, advertiser, merchant and creator accounts. 

    Any account may be deactivated for violating our policies. When a user is deactivated, all of their Pins and boards are also deactivated. That means that if you search for them or click on an old link to their profile, that profile won’t show up anymore. Their Pins won’t appear anywhere on Pinterest. And they won’t be able to access their own Pins or boards, either.

    Account deactivations, Oct–Dec 2020

    Policy Accounts
    Adult content 7,754
    Adult sexual services 494
    Civic misinformation 24
    Conspiracy theories 219
    Dangerous goods and activities 179
    Graphic violence and threats 11
    Harassment and criticism 816
    Hateful activities 2,487
    Medical misinformation 13
    Self-injury and harmful behavior 26
    Spam 3,115,438
    Account appeals and reinstatements

    If people believe their accounts have been deactivated by mistake, they can follow an appeals process to have their accounts reinstated. We review appeal requests and grant the appeal if we decide we made a mistake, or in some cases to give people a second chance to abide by our Community guidelines.

    Account appeals and reinstatements, Oct–Dec 2020

    Policy Appeals Reinstatements*
    Adult content 1,813 847
    Adult sexual services 38 2
    Civic misinformation 3 0
    Conspiracy theories 22 12
    Dangerous goods and activities 6 0
    Graphic violence and threats 7 3
    Harassment and criticism 24 13
    Hateful activities 31 8
    Medical misinformation 3 0
    Self-injury and harmful behavior 1 0
    Spam 99,839 64,777

    We also process appeals for deactivated Pins and boards, and expect to include that data in reports in the future.

    *Updated list on September 28, 2021 to reflect revised reinstatements figures.

    Law enforcement and government agency requests

    This section provides insight into the volume of information and deactivation requests received from law enforcement and government agencies. 

    Please note that this data encompasses requests from July to December 2020. For more information on how we respond to requests for account information, refer to our Law enforcement guidelines.

    Government Information Requests

    Pinterest receives legal requests from law enforcement and government agencies for Pinterest account information. We diligently review each request, and only produce data for those that meet the requirements of law and our policies. Our policy is to notify users of government requests for their information prior to disclosing, unless we are prohibited by law or in exceptional circumstances.

    United States Jan–June 2020

    Types Requests Some Information Produced Accounts   Accounts Notified**
    Subpoena 28 28 34 3
    Court Order  3 3 5 0
    Warrant 14 14 15 2
    Other* 1 1 1 0
    Total 46 46 55 5

    United States Oct–Dec 2020

    Types Requests Some information produced Accounts identified Accounts notified**
    Subpoena 41 28 45 4
    Court order 2 1 1 0
    Warrant 27 19 28 9
    Other* 6 3 5 0
    Total 76 51 79 13

    * Law enforcement requests such as wiretap orders, pen registers, trap and trace and emergency disclosure requests.
    ** The account owner was notified before production.

     

    International Jan–June 2020

    Country Types Requests Some Information Produced Accounts Accounts Notified**
    Canada Other* 1 1 1 0
    Total -- 1 1 1 0

    International Oct–Dec 2020*

    Country Requests Some information produced Accounts identified Accounts notified**
    Australia 1 0 1 0
    Belgium 1 0 0 0
    Brazil 1 0 1 0
    Germany 4 0 3 0
    India 3 0 1 0
    Switzerland 1 0 1 0
    United Kingdom 1 0 1 0
    Total 12 0 8 0

    * Law enforcement requests such as wiretap orders, pen registers, trap and trace, and emergency disclosure requests.
    ** The account owner was notified before production.

    National Security Requests*

    Time Period No. of Requests
    January to June 2020 0-249
    July to December 2020 0-249

    *National Security Request: Any national security letters and orders issued under the Foreign Intelligence Surveillance Act for user information

    Government Content Removal Requests

    We sometimes receive requests from government agencies to deactivate content on Pinterest that may be illegal in their country and/or a violation of our Community guidelines

    We review the requests to determine if the content identified violates our Community guidelines. Our teams take action on violations, ranging from deactivating the content globally to restricting access to the content within the relevant country if it violates local law but does not violate our policies. 

    Pinterest received a total of 4,078 requests from July to December 2020. We removed content for 3,684 of those requests for violating our Community guidelines and restricted the content on the remaining 394 requests.

    Jan–June 2020

    Country  Community Guidelines Violation* Legal Removals**
    Belgium 6 0
    Germany 1 28
    Kazakhstan 2 0
    Korea 76 1
    Russia 687 96
    Turkey 18 4
    Total 790 129

    Jul–Dec 2020

    Country Requests Community guidelines deactivation* Local law deactivation**
    India 14 0 14
    South Korea 2,314 2,048 266
    Russia 1,622 1,548 74
    Turkey 128 88 40
    Total 4,078 3,684 394

    * Content violated our Community guidelines and was removed from the platform.
    ** Content was reported by a government agency but did not violate our Community guidelines and was restricted from appearing only in the country where the request originated, based on local law.

    Child Safety

    Pinterest does not tolerate child sexual abuse material (CSAM). We have a strict no-tolerance policy for any content—imagery or text—that exploits or endangers minors. Detecting and deactivating this content is extremely important to us and we work closely with the National Center for Missing and Exploited Children (NCMEC) to combat this type of activity.

    Pinterest proactively identifies CSAM images and videos using matching tools including PhotoDNA and machine learning tools to detect additional possible CSAM. Our team of specialists is trained to identify and review CSAM, and was responsible for 1,794 CyberTipline reports to NCMEC from July to December 2020.

    Time Period Cybertip Reports
    January to June 2020 1,638
    July to December 2020 1,794

     

    Building toward a safer internet

    Creating the most positive space online doesn't happen by accident: It happens through proactive policy and product decisions. We have industry-leading positions on content safety that are informed by inputs and advice from outside experts, civil society and government. We also invest heavily in measures like machine learning technology to maintain a safe and positive space for the people on Pinterest. We’re proud of what we’re doing to keep Pinterest safe and to move the broader industry forward. 

    Let’s create a safer, more inspired internet, together.

    2019
    US Law Enforcement Information Requests

    Q1 January - March 2019

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 10 10 11 4
    Court order 0 0 0 0
    Warrant 9 9 9 3
    Total 19 19 20

    7

     

    Q2 April - June 2019

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 17 15 18 7
    Court order 0 0 0 0
    Warrant 11 10 12 3
    Total 28 25 30 10

     

    Q3 July - September 2019

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 11 11 17 2
    Court order 1 1 1 0
    Warrant 23 23 26 7
    Total 34 34 44

    9

     

    Q4 October - December 2019

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 7 7 7 0
    Court order 1 1 1 0
    Warrant 7 7 10 7
    Total 15 15 18 7

     

    Non-US Government Information Requests

    Time period Number of requests Some information produced Accounts Countries
    January - March 2019 0 0 0 n/a
    April - June 2019 0 0 0 n/a
    July - September 2019 0 0 0 n/a

    October - December 2019

    3 3 4

    United Kingdom, Brazil

     

    National Security Requests
    Time period Number of requests
    January - June 2019 0-249
    July - December 2019 0-249

     

    Government Content Removal Requests
    Type of violation // Time period Community Guidelines violation Legal removals Total Did not comply
    January - March 2019

    Germany (12)

    Russia (20)

    Turkey (1)

    Germany (11)

    Russia (45)

    Turkey (1)

    90 1
    April - June 2019

    Russia (56)

    Turkey (9)

    Germany (4)

    Russia (15)

    84 0
    July - September 2019

    India (12)

    Russia (168)

    Turkey (13)

    Germany (76)

    India (6)

    Russia (59)

    Turkey (1)

    335 0
    October - December 2019

    Australia (2)

    Germany (2)

    India (11)

    Russia (110)

    Germany (1)

    India (31)

    Russia (13)

    Turkey (1) 

    175 0
    2018
    US Law Enforcement Information Requests

    January - March 2018

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 6 4 4 3
    Court order 1 1 1 0
    Warrant 5 5 6 1
    Total 12 10 11

    4

    April - June 2018

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 7 4 5 3
    Court order 1 1 1 0
    Warrant 4 6 6 3
    Total 12 11 12 6

    July - September 2018

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 10 8 12 6

    Court order

    2 2 2 1
    Warrant 8 8 12 2
    Total 20 18 28 9

    October-December 2018

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 7 7 8 1
    Court order 1 1 1 0
    Warrant 5 5 5 1
    Total 13 13 14 2
    Non-US Government Information Requests
    Time period Number of requests Some information produced Accounts Countries
    January - March 2018 0 0 0 n/a
    April - June 2018 0 0 0 n/a
    July - September 2018 1 1 2 UK
    October - December 2018 0 0 0 n/a
    National Security Requests
    Time period Number of requests
    January - June 2018 0-249
    July - December 2018 0-249
    Government Content Removal Requests
    Type of violation // Time period Community Guidelines violation Legal removals Total Did not comply
    January - March 2018 India (1), Russia (21) Russia (4) 26 0
    April - June 2018 India (3), Korea (31), Malaysia (1), Pakistan (1), Russia (9), Turkey (5) Russia (3) 53 0
    July - September 2018 Russia (3), Turkey (5) Russia (5) 11 0
    October - December 2018 France (1), Russia (13), Turkey (4) Russia (44), Turkey (5) 67 0

     

     

    2017
    US Law Enforcement Information Requests

    January - March 2017

    Types Requests Some information produced Accounts Accounts notified*
    Subpoena 37 33 42 18
    Court Order 2 2 2 0

    Warrant

    39 39 42 22
    Total 78 74 86 40

    April - June 2017

    Types Requests Some Information Produced Accounts Accounts Notified*
    Subpoena 22 21 24 8
    Court Order 0 0 0 0

    Warrant

    20 19 20 5
    Total 42 40 44 13

    July - September 2017

    Types Requests Some Information Produced Accounts Accounts Notified*
    Subpoena 13 12 12 4
    Court Order 1 1 1 0

    Warrant

    9 9 11 1
    Total 23 22 24 5

    October - December 2017

    Types Requests Some Information Produced Accounts Accounts Notified*
    Subpoena 5 3 3 1
    Court Order 0 0 0 0

    Warrant

    7 7 7 0
    Total 12 10 10 1

    *Accounts notified means that the account owner was notified before production

    Non-US Government Information Requests
    Time Period Number of Requests Some Information Produced Accounts Countries
    January - March 2017 0 0 0 n/a
    April - June 2017 0 0 0 n/a
    July - September 2017 0 0 0 n/a
    October - December 2017 0 0 0 n/a
    National Security Requests
    Time Period Number of Requests
    January - June 2017 0-249
    July - December 2017 0-249
    Government Content Removal Requests
    Time Period Number of Requests Complied with Countries
    January - March 2017 72 72 Russia, South Korea
    April - June 2017 52 52 Russia
    July - September 2017 57 56 Russia, South Korea
    October - December 2017 23 23  

    *When Pinterest complies with a government request to remove content, we restrict that content from appearing only in the country where the request originated. That content will still be available to all other users.

    2016
    US Law Enforcement Information Requests

    January - March 2016

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 5 5 5 1
    Court Order 1 1 1 0

    Warrant

    0 0 0 0
    Total 6 6 6 1

    April - June 2016

    Types Requests Some Information Produced Accounts Accounts Notified
    Subpoena 8 6 13 2
    Court Order 1 1 12 0

    Warrant

    9 8 10 1
    Total 18 15 35 3

    July - September 2016

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 14 11 75 71
    Court Order 1 1 1 1

    Warrant

    5 5 8 5
    Total 20 17 84 77

    October - December 2016

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 10 10 12 4
    Court Order 1 1 2 0

    Warrant

    15 14 18 2
    Total 26 25 32 6
    Non-US Government Information Requests
    Time Period Number of Requests Some Information Produced Accounts Countries
    January - March 2016 0 0 0 n/a
    April - June 2016 0 0 0 n/a
    July - September 2016 0 0 0 n/a
    October - December 2016 0 0 0 n/a
    National Security Requests
    Time Period Number of Requests
    January 2016 - June 2016 0-249
    July 2016 - December 2016 0-249
    Government Content Removal Requests
    Time Period Number of Requests Complied with Countries
    January - March 2016 1 0 USA
    April - June 2016 1 0 USA
    July - September 2016 6 6 RUSSIA
    October - December 2016 1 1 RUSSIA
    2015
    US Law Enforcement Information Requests

    January - March 2015

    Types Requests Some information produced Accounts Accounts notified
    Subpoena 7 7 10 4
    Court Order 0 0 0 0

    Warrant

    2 2 3 0
    Total 9 9 13 4

    April - June 2015

    Types Requests Some Information Produced Accounts Accounts Notified
    Subpoena 9 8 11 7
    Court Order 4 3 6 0

    Warrant

    6 6 8 6
    Total 19 17 25 13

    July - September 2015

    Types Requests Some Information Produced Accounts Accounts Notified
    Subpoena 8 8 9 3
    Court Order 1 1 2 0

    Warrant

    2 2 2 2
    Total 11 11 13 5

    October - December 2015

    Types Requests Some Information Produced Accounts Accounts Notified
    Subpoena 6 6 6 2
    Court Order 0 0 0 0

    Warrant

    1 1 1 0
    Total 7 7 7 2
    Non-US Government Information Requests
    Time Period Number of Requests Some Information Produced Accounts Countries
    October - December 2015 1 0 1 Germany
    July - September 2015 0 0 0 n/a
    April - June 2015 0 0 0 n/a
    January  - March 2015 0 0 0 0
    National Security Requests
    Time Period Number of Requests
    July 2015 - December 2015 0-249
    January 2015 - June 2015 0-249
    Government Content Removal Requests
    Time Period Number of Requests Complied with Countries
    October - December 2015 1 1 Russia
    July - September 2015 3 3 Russia (3)
    April - June 2015 0 0 n/a
    January - March 2015 0 0 n/a
    2014
    January - March 2014

    Government Content Removal Requests: 0

    United States Law Enforcement Requests; Number of accounts: 7

    Warrant: 0 (0%); Protective Orders Obtained: 0

    Subpoena: 7 (100%); Protective Order Obtained: 4

  • Grand Jury: 5 (72% of subpoenas)

  • Administrative: 2 (28% of subpoenas)

  • National security: 0

    Other: 0

    Requesting party & compliance

    Federal: 2 (2 subpoenas); Compliance w/partial or full production: 100%

    State/Local: MN (2), FL (1), IL (1), CA (1); Compliance w/partial or full production: 100%

    United States Civil Requests: 0

    Rest of World Requests: Australia (1); Number of accounts: 1 ; Compliance w/partial or full production: 0%

    April - June 2014

    Government Content Removal Requests: 0

    United States Law Enforcement Requests; Number of accounts: 19

    Warrant: 3 (33%); Protective Orders Obtained: 2

    Subpoena: 6 (67%); Protective Order Obtained: 1

  • Administrative: 4 (67% of subpoenas)

  • Court: 2 (33% of subpoenas)

  • National security: 0

    Other: 0

    Requesting party & compliance

    Federal: 3 (2 warrants, 1 subpoena); Compliance w/partial or full production: 33%

    State/Local: WI (2), MO (1), OH (1), GA (1), CA (1); Compliance w/partial or full production: 83%

    United States Civil Requests: 0

    Rest of World Law Enforcement Requests: Canada (1); Number of accounts: 3 ; Compliance w/partial or full production: 0%

    July - Sept 2014

    Government Content Removal Requests: 0

    United States Law Enforcement Requests; Number of accounts: 15

    Warrant: 3 (43%); Protective Orders Obtained: 3

    Subpoena: 7 (57%); Protective Order Obtained: 4

  • Grand Jury: 4 (57% of subpoenas)

  • Administrative: 0 (0% of subpoenas)

  • Court: 3 (43% of subpoenas)

  • National security: 0

    Other: 0

    Requesting party & compliance

    Federal: 4 (1 warrant, 3 subpoenas); Compliance w/partial or full production: 100%

    State/Local: CA (1), PA (1), NJ (1), TX (1), AK (1), TN (1); Compliance w/partial or full production: 100%

    United States Civil Requests: 3 ; Compliance w/partial or full production: 66%

    Rest of World Requests: None; Number of accounts: 0 ; Compliance w/partial or full production: 0%

    Oct - Dec 2014

    Government Content Removal Requests: 0

    United States Law Enforcement Requests; Number of accounts: 19

    Warrant: 4 (30%); Protective Orders Obtained: 3

    Subpoena: 9 (60%); Protective Order Obtained: 6

  • Grand Jury: 5 (55 % of subpoenas)

  • Administrative: 1 (11% of subpoenas)

  • Court: 3 (33% of subpoenas)

  • National security: 0

    Other:0

    Requesting party & compliance

    Federal: 4 (0 warrants, 4 subpoena); Compliance w/partial or full production: 75%

    State/Local: MI (1), IL (1), MN (1), AZ (1), ID (1), WI (1), VI (1), PA (1), ND (1); Compliance w/partial or full production: 77%

    United States Civil Requests: 0

    Rest of World Requests: None; Number of accounts: 0 ; Compliance w/partial or full production: 0%

    2013
    July - September 2013

    Government Content Removal Requests: 0

    United States Law Enforcement Requests; Number of accounts: 7

    Warrant: 4 (57%); Protective Orders Obtained: 2

    Subpoena: 3 (43%); Protective Order Obtained: 1

  • Grand Jury: 1 (33% of subpoenas)
  • Court: 2 (66% of subpoenas)
  • Other: 0

    National security: 0

    Requesting party & compliance:

    Federal: 1; Compliance w/partial or full production: 100%

    State/Local: CA (2), FL (1), UT (1); NY (1) & WI (1); Compliance w/partial or full production: 83%

    United States Civil Requests: 0

    Rest of World Law Enforcement Requests:0

    October - December 2013

    Government Content Removal Requests: 0

    United States Law Enforcement Requests; Number of accounts: 5

    Warrant: 3 (60%); Protective Orders Obtained: 0

    Subpoena: 2 (40%); Protective Order Obtained: 0

  • Grand Jury: 1 (50% of subpoenas)
  • Court: 1 (50% of subpoenas)
  • Other: 0

    National security: 0

    Requesting party & compliance:

    Federal: 0

    State/Local: CA (2), FL (1), UT (1), OR (1); Compliance w/partial or full production: 100%

    United States Civil Requests: 1 (100%); Compliance w/partial or full production: 100%

    Rest of World Requests: 0

    End of Other articles Links
    Still need help? Contact us
    User feedback
    Was this article helpful?

    collection_fields

    How can we make this article better?