
Dutch file-sharing service WeTransfer is under fire after users spotted sweeping updates to its terms of service that appeared to let the company train AI models on their uploaded files.
The company has now removed the controversial language, but users remain outraged. Here’s what’s going on — and why it matters.
What did WeTransfer change?
WeTransfer users discovered this week that the service had updated its policy with a clause granting it a perpetual, royalty‑free license to use user‑uploaded content, including for “improving machine learning models that enhance content moderation.” The changes were due to come into effect on August 8.
That language was vague enough that many users — including children’s book writer Sarah McIntyre and comedian Matt Lieb — felt it opened the door for WeTransfer to use or even sell their files to train AI without permission or compensation.
How is this acceptable, @WeTransfer? You’re not a free service, I *pay* you to shift my big artwork files.
I DON’T pay you to have the right to use them to train AI or print, sell and distribute my artwork and set yourself up as a commercial rival to me, using my own work.😡 pic.twitter.com/OHPIjRGGOM
— Sarah McIntyre (@jabberworks) July 15, 2025
TNW City Coworking space – Where your best work happens
A workspace designed for growth, collaboration, and endless networking opportunities in the heart of tech.
How did WeTransfer respond?
On Tuesday afternoon, WeTransfer scrambled to douse the flames, saying in a press release that it doesn’t use user content to train AI, nor does it sell or share files with third parties. The company says it considered using AI to “improve content moderation” in the future, but that such a feature “hasn’t been built or deployed in practice.”
WeTransfer has also now amended its terms of service, removing any mentions of machine learning. The updated version states that users grant the company “a royalty-free license” to use their content for “operating, developing, and improving the service.”
But the damage to user trust may already be done.
Why are users so concerned?
WeChat joins a growing list of companies that have attracted criticism for training machine learning systems on user data. Adobe, Zoom, Slack, Dropboxand others have also recently walked back or clarified similar AI-related policies after public outcry. All these incidents tap into wider frustrations around copyright and consent in the AI age — and point to trust issues between users and tech firms.
WeTransfer has long marketed itself as a creative-friendly, privacy-conscious file-sharing service. So it’s perhaps unsurprising that the vague wording around AI and sweeping license rights felt like a betrayal to its users, particularly for artists and freelancers worried their work could be quietly fed into machine learning models without consent.
While WeTransder did clarify its terms, for many users of the service, the damage was already done. In replies to WeTransfer’s official announcement on X, some said that it looked like the service had tested the waters with broader AI permissions, got swift public backlash, and then quickly walked it back.
WeTransfer is unlikely to be the last tech firm caught up in this kind of controversy. As AI fever spreads, user data is becoming the new fuel.