Wikipedia has reached content licensing agreements with several AI giants, allowing its use in training large models.

ME News reported on January 17th (UTC+8) that the Wikimedia Foundation recently announced agreements with several AI companies, including Microsoft, Google, Amazon, and Meta, authorizing the use of Wikipedia content to train large-scale models through the Wikimedia Enterprise platform. New signatories also include Mistral AI and Perplexity. This move aims to address the trend of declining traffic due to AI summaries and to provide the foundation with sustainable revenue. Wikipedia is currently the only non-profit organization among the world's top ten websites, receiving 15 billion monthly visits and containing over 65 million articles. (Source: ME)