Mastodon Prohibits AI Model Training in Updated Terms
As concerns grow over how tech companies use public data to train AI models, Mastodon prohibits AI model training through a major update to its terms of service. Starting July 1, 2025, the decentralized social platform will explicitly ban any form of user data scraping for artificial intelligence development. This move comes on the heels of similar actions by X (formerly Twitter), Reddit, and other digital platforms seeking to protect user privacy from unregulated data extraction. For users wondering how Mastodon's stance affects their content and data security, the new policy provides a clearer, firmer stance against unauthorized usage.
New Terms Clarify Restrictions on AI Scraping and Automation
Mastodon’s updated terms directly call out prohibited behaviors, specifically the scraping of user data for AI model training and archival use. The platform now bans the use of any automated system — including bots, spiders, scrapers, or cheat utilities — to gather data from its servers. These policies aim to block large language model (LLM) developers and other data aggregators from extracting user content without permission. Notably, the only exceptions are common tools like search engines and browsers performing basic caching or human interactions on the site.
Limits Apply Only to Mastodon.social — Not the Entire Fediverse
While the update is a firm step forward, it’s important to understand its scope. These new rules apply solely to the Mastodon.social instance, which is the official server operated by Mastodon. Since Mastodon is part of the broader fediverse — a decentralized network of independently hosted servers — individual instances must adopt similar rules independently to ensure full protection. Without uniform terms across the network, there remains a loophole: other servers could still allow scraping unless they explicitly update their own policies.
Mastodon Raises Age Limit and Joins a Broader Industry Trend
Alongside its AI policy update, Mastodon is also increasing the minimum age for users. Previously set at 13 for U.S.-based users, the new age restriction is now globally set to 16. This aligns the platform with broader international data protection standards, including GDPR-related regulations in Europe. Mastodon’s decisions echo a larger industry-wide push, with platforms like OpenAI and The Browser Company also revising terms to limit unauthorized AI training. As the debate over ethical AI development continues, Mastodon’s actions highlight the growing importance of user consent and control over personal data online.
Post a Comment