The parent company of WordPress and Tumblr, Automattic, is reportedly in talks to sell training-related content from its platforms to artificial intelligence companies like MidJourney and OpenAI. This information was first reported by Gizmodo and was first linked to 404 Media. Automattic is emphasizing to consumers that they will always have the choice to refuse to have their data used to train artificial intelligence, even if the exact terms of the agreement are yet unknown.

According to the report, there exists an internal dispute inside Automattic around the incorporation of private material that was unintentionally taken from the internet for AI training, which goes against the company's planned protocols. Complicating matters, it has been revealed that commercial content that is not owned by Automattic—such as elements from an earlier Apple Music campaign—had entered the training dataset.

Furthermore, the story notes that a product manager at the firm has begun removing his images from Tumblr in response to the incident, presumably to prevent AI from using them for training. Automattic is going to provide a new feature that would allow customers to opt out of having their data shared, which is in line with the company's pledge.

The business described how this new option will "give users more control" over their material in a blog post. In the statement, the business also mentioned that it is actively working with certain AI entities to make sure that they are in line with community values including user control, opt-outs, and attribution. The business stated that all partnerships will respect opt-out options and that further efforts will be made to inform partners on a regular basis about those who choose not to participate. Additionally, the business stated that it hopes to make it easier for consumers to request future training and the removal of material from previous sources.