Patreon Blocks AI Training Bots in New Push to Protect Creator Content

Patreon has introduced stronger measures to prevent artificial intelligence companies from scraping creators’ content for model training, announcing a new partnership with internet infrastructure provider Cloudflare to block AI training bots actively. The move marks a shift from simply requesting that automated crawlers respect its policies to directly preventing unauthorised access to creator content.

The membership platform said its earlier protections, introduced in 2023, relied largely on robots.txt files, which instruct web crawlers on whether they are allowed to access certain parts of a website. However, Patreon said AI scraping techniques have become increasingly sophisticated, with some training bots ignoring those instructions altogether. As a result, the company has decided that passive requests are no longer sufficient to protect creators’ work.

Patreon has long benefited from the fact that much of its content sits behind a paywall, limiting access by automated crawlers. But newer features, including a redesigned Home Feed and Quips, its short-form social posting feature, expose more publicly accessible content that AI companies could potentially harvest for training large language models.

To address the issue, Patreon is extending its collaboration with Cloudflare by adopting the company’s AI Crawl Control technology. Rather than asking AI crawlers not to collect data, the platform will now actively identify and block bots that gather creators’ content for AI model training without permission. According to Patreon, the change is based on a simple principle: consent should not depend on whether an AI scraper chooses to follow the website’s instructions.

During testing of the new system, Patreon said weekly attempts by AI training bots to access its platform fell from thousands to zero after the new controls were enabled. The results suggest that many of the crawlers had previously ignored the platform’s robots.txt policies and continued scraping creator content despite explicit requests not to do so.

The company noted that not all automated bots will be blocked. Search engine crawlers and other indexing services that organise information and direct users back to Patreon will continue to be permitted because they help creators grow their audiences rather than collect material for AI training. The new policy is specifically aimed at bots designed to ingest content for artificial intelligence models.

The announcement comes as publishers, media companies and online creators increasingly question how AI developers collect data to train their models. Cloudflare has recently expanded its tools to help website operators manage AI crawlers, including the introduction of Pay Per Crawl. This marketplace allows publishers to charge AI companies for scraping their content. The company has also tightened its default policies for mixed-use crawlers that both index websites and collect data for AI training.

Patreon said the new protections are intended to give creators greater control over how their work is used as AI technologies become more widespread. The company believes creators should not have to choose between growing an online audience and maintaining ownership over the content they produce.

The latest changes reinforce Patreon’s broader position that creators deserve meaningful consent over AI training practices. By moving from voluntary compliance to active enforcement, the platform is seeking to establish stronger safeguards for creative work at a time when concerns over AI data collection continue to grow across the publishing and creator economy.