Do We Read the Terms of Use of Websites? On Paper Yes, In Practice No

When a popular file-sharing platform, such as WeTransfer, changes its terms of use, the reaction of most users boils down to one thing; click the ‘I agree’ box and move on. Because who actually reads the fine print when accepting terms of use?

However, someone did read the new terms of use of WeTransfer, which suggest that when users click ‘I accept’, they grant the platform a ‘lifetime, worldwide, non-exclusive, royalty-free, and transferable license with the right to sublicense’ for the use of content, its distribution, and the training of various AI models. All of this has triggered negative reactions from users who do not want their content to be at the mercy of a single company.

After receiving negative feedback, WeTransfer issued a statement on July 17 to clarify what the new terms actually mean. The statement explains that WeTransfer ‘does not use machine learning or any form of artificial intelligence to process content shared via WeTransfer’ and clarifies that ‘such a feature has not been built or used in practice, but was considered for the future’.

Since then, WeTransfer has removed any mention of machine learning or artificial intelligence from its terms of service, which now state: ‘In order to manage, provide, and improve our service and our technologies, we need to obtain certain rights related to content that is covered by intellectual property rights. By this, you grant us a royalty-free license to use your content for the purpose of managing, developing, and improving the Service, all in accordance with our Privacy and Cookie Policy.’

No One Reads It

WeTransfer concluded its statement with clarifications explaining that it does not use user content to train AI models and that ‘your content is always your content‘. It also promises that its terms are ‘in accordance with applicable privacy laws, including GDPR’ and that it does not ‘sell your content to third parties’.

Well, the uproar has passed, WeTransfer has implicitly apologized for the unclear terms, and we move on to new victories, but all of this has led us to wonder what about those who accept the terms of use of a platform without reading them, and how legally binding is just one click on a website? We posed this question to attorney Vlaho Hrdalo.

– The answer I would have to give as a lawyer is that they are legally binding, but I have a problem with that because theory is in serious discord with reality. And the reality is that internet users perceive it as just another box to check to finally access content that is – unlike the early internet of the late last century – wrapped in a force of marketing-legal decorative paper that is a hassle to unwrap every time – says Hrdalo.

In other words, a contractual obligation exists, but awareness of it generally does not. That is why, as Hrdalo says, ‘absolutely no one’ reads them, except when it comes to B2B terms of use, where minimal effort is made to see what is being agreed to.

However, users cannot escape with the ‘I didn’t know’ argument because, as Hrdalo states, we do know that we are agreeing to something.

– The reason we don’t care much about the content of the general terms is that generally, due to these blind acceptances, we do not face any special consequences. However, the basic division of the ways of consent is into implicit and explicit, or browsewrap and clickwrap. While the former does not require a specific action from the user, the latter does, which guarantees greater success in court for the one who set those terms. A well-known case is the one from last year where a user wanted to sue Uber in court for damages due to an accident caused by their vehicle, but the court directed him to arbitration, which the user unknowingly agreed to by accepting Uber’s general terms – explains Hrdalo.

AI Raises New Questions

Another sensitive area is copyright. Platforms like WeTransfer often incorporate clauses about licenses for the content that users upload into their terms. This means that a private photo or document, without the author’s knowledge, can become subject to the platform’s rights, no matter how loudly they proclaim ‘your content is yours’.

– Granting a license in my opinion should not be possible in such a way because in that case, the user is often not aware of what they are uploading. In my opinion, disputing that a license was granted through a tacit act such as checking a box could be successful. I note that courts generally still do not think so because cases often come before them where the licensor is someone for whom this is a business. Here we are talking about a situation where a private individual uploads their photo to WeTransfer, or similar platforms, and then they claim copyright over it. This will have to be redefined – explains Hrdalo.

This shows that the legal framework has still not fully adapted to everyday digital practice, where content is often created and shared outside of a professional context. The additional challenge is posed by artificial intelligence.

– I believe that the use of AI will lead to terms being changed much more frequently. This will, in turn, lead to the opening of new questions. Because even if we agree that the user has unreservedly accepted the general terms by clicking the designated box, how often do we expect them to review them? It is not realistic to regularly check if something has changed in the terms on a site you visit every day. An internet user is not a professional lawyer, nor should they be – concludes Hrdalo.

Therefore, it should be concluded that the changes made by WeTransfer and similar services show the gap between legal theory and digital practice. On paper, users are informed and aware of what they are agreeing to. In reality, the decision boils down to a routine click, without reading and understanding.