WeTransfer was forced to respond this week after changes to its terms of service (TOS) triggered major backlash from users who believed the new language granted the service access to users’ files to train AI.
“We don’t use machine learning or any form of AI to process content shared via WeTransfer, nor do we sell content or data to any third parties,” a WeTransfer spokesperson told BBC News on Tuesday.
WeTransfer clarified this after users noticed recent changes to its TOS page, which initially said the following policy would go into effect in August (via Wayback Machine on July 14, 2025).
You hereby grant us a perpetual, worldwide, non-exclusive, royalty-free, transferable, sub-licensable license to use your Content for the purposes of operating, developing, commercializing, and improving the Service or new technologies or services, including to improve performance of machine learning models that enhance our content moderation process, in accordance with the Privacy & Cookie Policy.
The language seemed to imply that WeTransfer could use data and files from users to train AI models, either their own or that of a third party. Outrage from users swiftly followed, many of whom are independent artists who use WeTransfer to send large files like film footage or music.
Mashable Light Speed
This Tweet is currently unavailable. It might be loading or has been removed.
Users took to social media to call attention to the change, with some vowing to switch to another service. The issue of users’ content being used as AI training data is a contentious one that has become increasingly widespread as companies look to develop their own AI models and features; particularly since these tools that can automate creative work and have already impacted job markets. Users are wary of freshly-updated terms of service, since it could mean signing their data away to AI models and automating themselves right out of a job.
Similar confusions happened with other platforms, like CapCut’s policy update, which sounded alarming but is actually pretty standard. And Adobe had to clarify its policy changes last year, when the update made it sound like it was using creators’ content to train its Firefly model without permission. However, companies like Google and Meta rely on user data to train their models, following the whole “if something free, you’re the product” doctrine.
This Tweet is currently unavailable. It might be loading or has been removed.
This Tweet is currently unavailable. It might be loading or has been removed.
But WeTransfer has changed the language in the content section of the policy after acknowledging to BBC News that the previous update “may have caused confusion for our customers.” The company further clarified to the outlet that the original language was intended to “include the possibility of using AI to improve content moderation” for the purposes of identifying harmful content.
The section now reads:
You hereby grant us a royalty-free license to use your Content for the purposes of operating, developing, and improving the Service, all in accordance with our Privacy & Cookie Policy.
In both versions of WeTransfer’s TOS, the company states that the users own and retains all the rights to their work. That should clarify and confusion that WeTransfer was trying to take ownership of the work. Although WeTransfer does have a license to users’ files, it’s for the ostensive purpose of content moderation — not to train AI models.
Topics
Artificial Intelligence
Privacy
