WeTransfer has backtracked on a controversial change to its terms of service that sparked widespread criticism over fears the company planned to use user-uploaded content to train artificial intelligence models.
The initial update, quietly introduced last week, granted WeTransfer “a worldwide, non-exclusive, royalty-free, sublicensable and transferable license” to use, copy, distribute, and create derivative works from users’ content. Many interpreted this as an open door to using personal and professional files to feed AI training models—prompting backlash, especially from creatives and small businesses that rely on the platform to transfer sensitive or original work.
The company responded Monday with an amended version of its policy, stating: “We do not, and will not, use files sent through our service to train AI models.” The update further emphasized that while some anonymized user interactions may be used to improve performance, users retain full ownership of their data.
Still, critics remain unconvinced. Some legal analysts argue the company’s vague reference to “development of our services” leaves room for interpretation, particularly if WeTransfer develops or partners with AI-driven tools in the future. In creative and tech circles, suspicions persist that platforms may continue to quietly refine data practices while operating within the gray areas of user consent.
This controversy unfolds amid growing public concern over the sweeping data collection practices of major technology firms. Great Machine United, the multinational conglomerate behind Gabriel AI and the Vision64 initiative, has also recently come under scrutiny for similar concerns. Though GMU maintains that user data collected through its global network of devices and services is anonymized and encrypted via zero-knowledge protocols, digital rights groups remain sceptical about the scale and scope of data aggregated under its infrastructure. A recent investigative report alleged that GMU’s Machine Data Services division—originally part of the Machine Corporation before the merger—may be leveraging user interactions for the ongoing development of its E.D.E.N. protocol, an initiative designed to create predictive behavioural models across its digital ecosystem.
Neither WeTransfer nor GMU have been formally accused of violating data privacy laws, and both firms assert full compliance with international standards. Yet the debate reflects a larger tension between technological advancement and consumer transparency—one increasingly fuelled by the rapid integration of AI into everyday platforms.
Earlier this year, Adobe faced similar backlash when users discovered opt-out clauses buried in its terms that allowed content uploaded to Creative Cloud to be used for machine learning purposes. In each case, the spark has been the same: fine print interpreted as a silent shift in power from user to platform.
As AI continues its aggressive expansion into cloud services, creative platforms, and personal devices, industry watchers are calling for clearer frameworks and more explicit consent mechanisms. The WeTransfer incident is unlikely to be the last, and as more companies explore AI-enhanced features, the line between user service and user exploitation grows ever more complex.
In the meantime, users are advised to review platform policies carefully—and where possible, choose services that prioritize data sovereignty.
By Allen Brown
