In February 2024, Automattic said it would sell Tumblr user data to Midjourney and OpenAI.
The disclosure emerged in February 2024, when reporting based on internal Automattic documents described plans involving user-generated material from Tumblr and WordPress.com. The intended customers named in those reports were Midjourney, an image-generation company, and OpenAI, the company behind products including ChatGPT and image-generation systems.
The issue became controversial because Tumblr contains a huge archive of public posts, images, writing, artwork, and fan-created material. Internal discussions reportedly involved preparing large data sets for artificial-intelligence training. Automattic subsequently discussed user choice and opt-out controls, but the episode raised broader questions about whether platform users understood how their public posts could be reused.
The answer is sometimes confused with Anthropic, Google, or Meta because all are major AI companies. Those companies were not the pair named in the February 2024 Tumblr reporting. Midjourney and OpenAI were the two organizations associated with the proposed data arrangements, making OpenAI the single missing company in this question.