Mastodon Revises Terms to Ban AI Model Training

Social networks are tightening their terms of service to combat the unauthorized use of their data for AI training. Following X’s prohibition against AI model training by external parties, Mastodon has introduced similar restrictions. The updated terms explicitly prevent the scraping of user data for purposes like training large language models (LLMs). Effective July 1, Mastodon’s new rules discourage the use of automated systems, such as bots or scrapers, to access user data except for standard web searching and local caching. However, these restrictions apply only to the Mastodon.social server, meaning data could still be harvested from other instances in the fediverse that lack such prohibitions. Other platforms, including OpenAI and Reddit, have implemented similar clauses. Additionally, Mastodon is raising its global user age limit from 13 to 16.

Source link