Why Tumblr Banned Erotic Content — and What It Cost Everyone

What Tumblr Was

To understand what was lost, you have to understand what Tumblr was at its peak — roughly 2010 to 2018 — before the ban and the long decline that followed.

Tumblr was a microblogging platform built around the reblog: a system that allowed users to share other users' posts to their own feeds, building networks of taste and curation that had no equivalent anywhere on the internet. At its best, a well-curated Tumblr feed was an education in the specific aesthetic sensibility of the person who had built it — a window into exactly what they found beautiful, funny, erotic, politically urgent, or worth preserving.

For gay men specifically, Tumblr was something close to irreplaceable. It was the platform where gay erotic photography — serious photography, made with genuine craft and genuine desire — could be shared, curated, and appreciated without being buried under the industrial volume of mainstream pornographic platforms or forbidden entirely by the sanitised policies of mainstream social networks. The reblog culture meant that good images spread. Curators with good eyes built audiences. Photographers found viewers who appreciated their work for what it was.

The platform at its peak hosted work by serious photographers alongside amateur images, criticism alongside fantasy, political content alongside erotic content — all in the same feed, mixed according to the curator's judgement rather than the algorithm's. It was, in its way, the closest thing the internet had produced to the experience of a well-stocked, sexually honest, culturally engaged gay magazine: something made by gay men for gay men, without corporate interference or advertiser anxiety.

By 2018, Tumblr had around 521 million monthly visits. A significant portion of those visits were driven by adult content. The platform knew this. It had always known this.

Why the Ban Happened

The official reason was vague. Tumblr CEO Jeff D'Onofrio posted a statement announcing the ban and describing it as laying "the foundation for a better, more positive Tumblr." He did not explain what had been wrong with the existing Tumblr, or why the adult content that had been central to the platform's culture and traffic for a decade was now incompatible with its future.

The actual reason was simpler and less flattering. On November 20, 2018, Tumblr was removed from Apple's App Store after child pornography was discovered on the platform. The removal was a crisis. A former Tumblr iOS engineer described what followed: the app was rejected repeatedly by Apple until the adult content ban was fully enacted. While Tumblr was actively working on the ban, they were asking Apple for guidance on what would meet approval, and the answers they received were either vague or unhelpful. Tumblr had to keep re-submitting with a half-baked content-finding algorithm until the app finally looked clean enough for Apple.

There were compounding pressures. SESTA/FOSTA, signed into law in April 2018, made platforms liable for content tied to prostitution, meaning companies faced a choice between moderating the age, content, and consent of everyone publishing to their platforms, or risking criminal lawsuits. Companies — especially large ones — are risk-averse, and when you're being cut off from potential customers and could be liable for things related to sex, it is much easier to institute a blanket ban. Tumblr also faced restrictions in Indonesia and pressure from payment processors who were tightening their policies around adult content across the industry.

The child pornography issue was real and serious, and no one is arguing that Tumblr should have retained illegal content. But thousands of sites manage to effectively moderate to keep their platforms free of child pornography while allowing adult content more broadly. Tumblr took the unique step that many others haven't by completely banning any and all adult content, as opposed to just trying to increase their screening or security protocols.

The ban was not a proportionate response to the problem. It was a panicked, maximalist solution that sacrificed an entire cultural ecosystem to satisfy a corporate gatekeeper's content policies.

The Algorithm That Couldn't Tell Art from a Tyre

The implementation of the ban was, on top of everything else, a technical disaster. Tumblr deployed an automated moderation system to flag and remove adult content, and the system was, charitably, not ready for the task.

Sarah Burstein, a law professor with a Tumblr account that shared design patents, found a significant number of her posts being flagged. The automated filter objected to a heart-shaped necklace, a boot-scrubbing design, LED jeans, troll socks, a Louis Vuitton bag, some boxes, a tyre, a hanger, a flamingo pool float, shoes, pillows, and more — almost exclusively black-and-white line drawings and diagrams.

A crowdsourced list of banned tags on Tumblr's iOS app included, alongside the expected adult content terms, tags related to stimming (a coping practice common among autistic people), along with tags like "depression," "PTSD," and "bipolar." Even the numbers 69 and 420 were banned.

The artists and photographers whose work was the platform's most valuable cultural content — people who had spent years building audiences for serious erotic photography — watched their archives disappear behind flags or vanish entirely. The Electronic Frontier Foundation noted that Tumblr had insisted the policy struck a balance and would continue to foster diversity of expression for people discussing art, sex positivity, relationships, sexuality, and personal journeys — while deploying an automated system that was incapable of the nuance the policy claimed to embody.

The irony was complete: in attempting to remove illegal content, Tumblr's system targeted the thoughtful, the artistic, and the carefully made — exactly the content that had given the platform its cultural value.

What It Cost: The Numbers

The commercial consequences were immediate and severe. The platform reacted to the CSAM crisis by banning all adult content, resulting in a near-immediate 29% monthly traffic decrease. Monthly visits fell from approximately 521 million in December 2018 to around 370 million by February 2019 — a loss of 150 million visits in under two months.

The decline did not stop there. Traffic estimates for subsequent years tell a consistent story of slow, continuous loss: approximately 318 million visits in 2020, 327 million in 2021, 270 million in 2022, and around 213 million in 2023. Some recent analyses suggest a partial stabilisation, with around 135 million monthly active users in the mid-2020s, but this represents a much smaller platform than existed before December 17, 2018, and one that has never come close to recovering its pre-ban scale.

Roughly a fifth of users left in the immediate aftermath. App downloads for new users declined. The tech-savvy, early-adopter audience that had made Tumblr a cultural bellwether dispersed to platforms that tolerated their interests or to corners of the internet that were harder to find.

D'Onofrio's "better, more positive Tumblr" turned out to be a smaller, quieter, culturally diminished Tumblr. The foundation it laid was for irrelevance.

What It Cost: The Culture

The commercial numbers are only part of the story. The cultural cost was harder to quantify and more lasting.

The gay erotic photography community that had built itself on Tumblr over a decade did not simply relocate. It dispersed. The curators, photographers, and viewers who had found each other through the reblog system had no equivalent platform to move to — nothing that offered the same combination of visual focus, community infrastructure, and tolerance for adult content. Twitter tolerated adult content but was a firehose of everything else; Reddit had communities but lacked visual sophistication; dedicated adult platforms existed but had no culture of curation or aesthetic seriousness.

In a move aimed to increase the site's security, Tumblr ended up destroying a safe area for its most vulnerable user base — not just in terms of pornography consumption, but as a safe space for young LGBTQ people looking for allies online. The platform had been, for many young gay men in places where their identity was unwelcome, the first space where they had found images of themselves — images that reflected their desires honestly, without shame, made with genuine care. That space disappeared overnight.

The archive of images, curations, and communities that had been built over a decade was not preserved. It was not migrated. It was flagged, hidden, and in many cases permanently lost — deleted either by the moderation system or by creators who, confronted with the automated flags and the impossibility of the new system, simply gave up and went offline.

The Lesson Tumblr Didn't Intend to Teach

The Tumblr ban demonstrated, with unusual clarity, what happens when a platform that has grown on the back of a particular community abandons that community under corporate pressure. The lesson was not subtle: the community does not forgive, the traffic does not return, and the platform that survives the decision is a shadow of the one that made it.

It also demonstrated the danger of depending on a single platform — particularly one owned by a corporation with its own business interests — for cultural infrastructure that matters. Tumblr's gay erotic photography community had no backup, no alternative architecture, no redundancy. When the platform changed its policies, everything built on it was vulnerable.

The ban made the need for a purpose-built alternative undeniable. Not a platform that tolerates gay erotic photography among its many other functions, not a platform that can be pressured by an app store into destroying its own content overnight, but a platform built specifically for this purpose, with this audience in mind, with the architectural commitments that make the content's safety structural rather than incidental.

That is what [BoysDo](https://boysdo.com) is. The platform Tumblr should have remained. The gallery that didn't close.

A Note on What Tumblr Got Wrong

Tumblr's decision was wrong not because adult content moderation is unnecessary — it isn't, and the platform's failure to prevent illegal content from appearing on its servers was a genuine failure — but because the response was disproportionate, blunt, and made without any apparent consideration for the cultural value of what was being destroyed.

A serious, properly resourced content moderation system could have distinguished between illegal material and legal adult content, between exploitative content and artistic erotic photography, between the content that had caused the crisis and the content that had given the platform its value. The technology existed, if imperfectly. The will to build it, apparently, did not.

The decision to deploy a blanket ban rather than invest in proper moderation was, ultimately, a statement about how much the platform valued its adult content community. The community heard the statement clearly. They left, as communities do when they understand they are not wanted.

The loss is ongoing. The archive is gone. The community is scattered. And the platform that sacrificed them has been declining, slowly and consistently, ever since.

Part of the BoysDo editorial series. Explore the platform at [boysdo.com](https://boysdo.com).