AI & LLMs

Large language models, like ChatGPT, is a type of "artificial intelligence" program designed to understand and generate human-like text by processing vast amounts of written language. While the creation and use of large language models are not inherently unethical or illegal, there are many important questions about how companies obtain their data to train their models and how that data is used.

Data Training

Except where necessary to provide core Mastodon services such as indexing content to return search results via our web interface or API, we do not, and will not, train any data models using vmst.io or federated user data. We do not, and will not, provide access for third-parties to use vmst.io or federated user data for any training purposes.

Prohibited Uses

Our Terms of Service prohibit crawling the content of our site for AI training purposes.

vmst.io has taken some steps to limit organizations who build these models from accessing your public post data. In some cases, this involves directly blocking the accessibility of third-parties to our site through firewalls, but in most cases, this is done by requesting that they not index our site through user agent flags in our robots.txt file.