6

In a click-wrap scenario, how does one prove, legally, that a user actually clicked or checked the "i accept" or equivalent on a given site.

e.g. what if the website owner doesn't track clicks, track TOS versions, etc. What if the whole user flow changed since the alleged acceptance of TOS, etc.

What do the courts look for to determine whether the user did in fact take the action which confirms their agreement with the TOS, and what those TOS actually were at the time of the issue in question? ...and, with which party does the burden of proof rest and what constitutes proof?

2
  • The site owner might need to track such clicks in a database of some sort to be fully protected if a dispute arises. Jun 11, 2019 at 21:43
  • As an example, the user may have used their browser's Developer Tools to alter a page's DOM and bypass the ToS prompt. Feb 3, 2021 at 21:14

3 Answers 3

8

The website owner brings in an expert programmer who testifies that the user cannot have gotten to a certain part of the site (or download, etc.) without having clicked to accept the terms of service, and that this document they're holding is a true and correct copy of the terms of service as of that date. That's evidence in favor of the site, and an adverse party has to have stronger evidence in order to overcome it. If the person didn't save a copy of the terms themselves, they'll have a hard time on this.

Then the other party's attorney tries to discredit the programmer by asking questions like "how do you know there are no bugs in the software which could have allowed somebody to reach this without agreeing to the terms of service" etc. Apparently, some sites don't require users to click indicating agreement.

If the company has significantly changed the site, terms of service, etc. since the time the user registered, and doesn't keep any copies of old versions around, and admits this, they'll have a hard time enforcing an agreement (as they can't produce a copy of it). If the user kept a copy, the user might be able to present that.

It's up to the finder of fact to decide what to believe and how much weight to give the various witnesses' testimony.

10
  • OK, perfect... to clarify: If the website owner is to be protected, he (at minimum) should do the following: (1) track versions of the TOS over time including the content and change date/time (2) track which users accepted the TOS, recording the version accepted, date accepted, (3) fully document the software QA/testing process, specifically for features supporting the reading, acceptance/denial of, and logging of the TOS, and (4) retain server logs/web analytics logs which may show the clickstream of the user when they accepted/denied the terms of service. Does that make sense?
    – GWR
    Nov 10, 2015 at 11:56
  • 1
    It would seem that ideally, if the website owner had a way to postmark each new version of the TOS, this would be the best way to prove the content of the TOS at a given time, but how could one prove the record of a users acceptance or denial of such terms is a true record (and not a fake record inserted by an unscrupulous website owner)?
    – GWR
    Nov 10, 2015 at 11:56
  • Yes, those are records the website owner should be keeping - and any business should have a full copy of their terms of service documents with dates they used those. Archive services can be of help to show the particular content of a particular URL on a particular day. For example, you can enter a URL in the lower right corner on archive.org/web to archive it as of that day, assuming that the website allows crawlers. You can also adjust your site to make it easier for crawlers and archivers.
    – WBT
    Nov 10, 2015 at 16:18
  • 2
    @lol "Everything is recorded down to the keystrokes..." I don't think that's accurate, though if an opposing party were able to get access to the full git history (which seems unlikely) they could reconstruct at least most of the site as of the day they allegedly accepted the ToS, and then (if applicable) demonstrate a feasible path to sign up / create an account / etc. without indicating acceptance.
    – WBT
    Jun 27, 2016 at 20:56
  • 1
    I think we're getting too fixated upon the actual keystrokes? The exact keystrokes are irrelevant -- I'm saying that due to commit date stamping you can see the exact state of the codebase at any given time. Such fine-grained level of granularity is perfect to meet any evidentiary standards. But now that you mention it I must raise git history tampering: just because it's in a git history, doesn't mean it hasn't been tampered with (and force pushed to master, and altered on any other pulled copies - lol!)
    – lol
    Sep 30, 2016 at 13:48
3

WBT gives a great perspective from the legal side!

From the website owner's side I can add the following as I had to go through setting up the legal part of my software as a service platform recently:

  1. I keep a copy of all my legal documents (term of service, privacy policy, cookie policy) along with a timestamp of when they were published. I automated the process of archiving new revisions but you could very well do this by hand.

  2. When a new user signs up I store a timestamp of when the user clicked the "Sign up" button. Btw, my server verifies that the user checked "I agree..." or otherwise cancels the sign up.

  3. If I get into the unfortunate situation that I have to produce the legal documents in court I can compare the timestamps of the legal document history with the timestamp in the sign up record and then pick the revision of my legal documents the user agreed to back then.

-5

You would have to click OK to be let into the site right? So if you used the site you clicked ok.

1
  • 1
    The technical term for that is "begging the question";if you accept that nobody can access the site without clicking OK then everybody who accessed the site clicked OK. As indicated in the question, not everybody always accepts that. Jun 12, 2019 at 20:26

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .