This week, WeTransfer responded to worries after revisions to its terms of service (TOS) triggered backlash from users who were concerned the new terms would permit the service to utilize users’ files for AI training purposes.
“We do not engage machine learning or any type of AI to process content shared through WeTransfer, nor do we sell any content or data to third parties,” a WeTransfer representative informed BBC News on Tuesday.
WeTransfer provided this clarification after users observed modifications to its TOS page, which initially outlined a policy effective in August (as seen 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 enhancing the Service or new technologies or services, including to improve the performance of machine learning models that bolster our content moderation process, in accordance with the Privacy & Cookie Policy.”
The wording indicated that WeTransfer might leverage user data and files to train AI models, provoking user outrage, particularly among independent artists utilizing WeTransfer for large files such as film footage or music.
Users took to social media to showcase the modification, with some vowing to transition to alternative services. The issue concerning the use of users’ content as AI training data is contentious, as companies create AI models and features that could automate creative tasks and influence job markets. Users are wary of updated terms of service, fearing their data could be repurposed for AI models, potentially impacting their employment.
Similar uncertainty arose with other platforms, such as CapCut’s policy revision, which appeared concerning but was standard. Adobe also needed to clarify its policy changes last year, as it seemed to utilize creators’ content to train its Firefly model without authorization. However, businesses like Google and Meta depend on user data to train their models, adhering to the doctrine of “if something is free, you’re the product.”
WeTransfer has modified the wording in the content section of the policy after acknowledging to BBC News that the previous revision “may have created confusion for our customers.” The company explained that the original wording aimed to “include the possibility of using AI to enhance content moderation” for identifying harmful content.
The section now states:
“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 iterations of WeTransfer’s TOS, the company asserts that users maintain and retain all rights to their work, clarifying any ambiguity about WeTransfer gaining ownership. Although WeTransfer holds a license to users’ files, it’s specifically for content moderation purposes, not for AI training.