
The decentralized social networking platform Mastodon has recently issued an email to its users, notifying them of an update to its Terms of Service. The most significant change introduced is a formal prohibition against the unauthorized scraping of user data.
The revised terms will take effect on July 1, 2025. From that date, AI companies that continue to extract data without explicit permission may face legal action from Mastodon. This move follows similar policy updates from other platforms such as X (formerly Twitter) and Reddit, which have also barred the use of their data for AI model training.
In its notification email, Mastodon stated:
“We explicitly prohibit the scraping of user data for unauthorized purposes, e.g. archival or large language model (LLM) training. We want to make it clear that training LLMs on the data of Mastodon users on our instances is not permitted.”
The updated terms include the following clause:
“Use, launch, develop, or distribute any automated system, including without limitation, any spider, robot, cheat utility, scraper, offline reader, or any data mining or similar data gathering extraction tools to access the Instance, except in each case as may be the result of standard search engine or Internet browser and local caching or for human review and interaction with Content on the Instance.”
It is important to note that these terms apply solely to Mastodon’s officially maintained server, Mastodon.Social, which is but one instance within the broader federated Fediverse. If other servers within the Fediverse do not explicitly prohibit data scraping in their own terms, AI companies may still legally collect data from them for model training.
However, it is likely that other server operators may follow the lead of Mastodon’s official instance by adopting similar terms to restrict AI companies from harvesting data. For social platforms, user data represents an invaluable asset, and the unauthorized use of this data for AI training constitutes a significant loss.
Related Posts:
- Mastodon Alert: CVE-2024-23832 Unlocks Account Takeover Threat
- PoC Published for Critical Mastodon Vulnerability – CVE-2024-23832 (CVSS 9.8)
- CVE-2023-36460: Mastodon Patches Critical RCE Security Vulnerability
- Mozilla’s New Terms: Firefox to Collect User Data