Patreon Blocks Crawlers From Stealing Creators' Work for AI Training

Patreon dropped a blunt message this week, delivered in classic Jack Conte fashion: tech companies that scrape creators' work to train AI models can, in his words, stay the fuck off the platform. The Patreon founder and CEO announced on Instagram that the company has partnered with Cloudflare to block AI training crawlers from accessing content published on Patreon, framing it less as a policy update and more as a declaration of war on uncompensated scraping.

The blocking happens at the network level, according to Conte, and it's already live across every post published on Patreon. Drew Rowny, Patreon's SVP of Product, put it in less colorful but equally pointed terms in a Cloudflare press release, saying that most of the internet forces creators to accept AI training on their work just to build an audience. Patreon, he said, wants creators to grow their reach without surrendering control over how that work gets used — while still letting through the crawlers that help people get discovered via search.

This isn't Cloudflare acting alone for one client. The infrastructure company said last year it would start blocking AI crawlers from pulling content without site owners' permission or payment, as a default setting. Earlier this month it rolled out finer-grained controls, letting site owners decide separately whether to allow search crawlers, agent crawlers, or training crawlers. And starting in September, any new domain onboarding to Cloudflare will have training and agent bots blocked by default on ad-supported pages, with search crawlers still let in.

Patreon's move rides on that infrastructure, specifically a feature called Crawl Control, which a Cloudflare spokesperson confirmed Patreon recently switched on for its users. Beehiiv, the newsletter platform, has enabled the same tool recently, letting its users choose which AI models to allow or block.

Conte has been building toward this for a while. Back in May, he posted a 43-minute video laying out what he called the three things creators are owed and currently don't get from AI companies: consent to opt out of having their work used as training data, credit when a model essentially copies their style, and compensation when that happens. His answer to whether creators currently get any of that was, in his own words, a big fat no.

The timing is worth sitting with, though. Patreon still permits AI-generated content under its terms, and the platform has had its own reckoning with what gets monetized on it — 404 Media reported in 2024 that creators of nonconsensual sexual images and videos were making money there. Patreon has since tightened its rules, allowing illustrated AI depictions of people but requiring documented consent for hyperrealistic ones.