In X Corp v. Bright Data LTD, X Corp, formerly known as Twitter, sought to put an end to the practice of “scraping” by a company called Bright Data. (A scraper requests and retrieves information without manually browsing pages, often for tailored reasons.) The Northern District of California dismissed the case, finding that X failed to state a claim.
There were two ways in which X Corp accused Bright Data of violating X’s rights:
- •X Corp alleged that Bright Data improperly accessed the X service and assisted others to improperly access the service.
- •X Corp alleged that Bright Data improperly scraped and sold data and assisted others to do that.
The court said the first one didn’t work because X never said that Bright Data used an X account to scrape the data. That’s because, in fact, no one has to be logged in as an X account holder in order to scrape data. So, from the get-go, the breach of contract claims were disingenuous. Or, at least, worthy of dismissal.
X’s Terms of Use (TOS) starts out by stating, in plain language, that all users own their own content. However, in other places, the TOS says that any reproduction must be made using the interfaces that X provides. Furthermore, the TOS bans X users from scraping the data. But X did not suggest the data being scraped in these cases was accessible only to logged-in X users.
As an alternative to breach of contract, X claimed Bright Data engaged in tortious conduct: using elaborate technical measures to evade X’s anti-scraping technology. X Corp also tried to say that this amounted to trespassing on its data, but courts have previously held — although there’s some appropriate use of this ancient idea — that trespassing requires preventing the owner of the property from accessing or using the property. That was not even alleged by X; there’s no allegation that X Corp didn’t have access to the data that Bright Corp was scraping.
Now, the court is careful to point out that this doesn’t mean that someone in another case might be able to allege a trespass to chattel claim. For example, a denial-of-service attack could prove to be a viable claim because it’s preventing the owner from accessing and using its data. It’s also important to note that the court didn’t admonish X Corp for trying to stop Bright Data and others from scraping. However, in this case, essentially harmless interference did not constitute trespass to chattels.
A claim for breach of California’s Unlawful and Fraudulent Business Acts said that Bright Data deceived X Corp into providing access to its data, but this, too, was disingenuous. As the court noted earlier, one doesn’t need a registered account to access X and make requests for information. Therefore, the defendant didn’t misrepresent itself by posing as a legitimate X user but rather acted as a legitimate non-user who wanted to get information. So, the misrepresentation claim also failed.
Bright Data was not imitating anyone, but X Corp seized on the idea that Bright Data didn’t allow itself to be identified by any kind of particular IP address. There was no fraud there because anyone can have an IP address that changes.
The court said there could be a legitimate case of someone commiting fraud on a social media company who is found out and banned from the app — but then disguised themselves as a different user and went back to violating the terms. But that’s not what happened here.
The tortious interference with the contract claim also failed because there were no damages in this case. On the one hand, Bright Data did provide tools to X Corp’s account holders that enabled them to scrape data in violation of their contract. On the other hand, X Corp never showed any way in which it was harmed by data being scraped.
Bright Data itself acknowledged that it registered and agreed to the terms of X Corp. But having agreed to the terms doesn’t mean that there’s a violation or a breach of contract where X Corp. couldn’t allege suffering any damages. All X could do is allude to the diminished capacity caused by having too many scrapers come in and overload their servers, but they never demonstrated that ever happened.
Eventually, the court arrived at the interesting copyright problem, in which X Corp says it has the right, and royalty-free license, to use the content of its account holders — but it wants to stop others, and exclude others, from scraping (using, copying, or reproducing) the data of X users’ content. So, on the one hand, X Corp disclaims owning any of the X users’ content; on the other hand, it wants to be the exclusive distributor of that content as long as it gets paid for reselling it to other companies.
If this claim were viable, the court said, X Corp would entrench its own private copyright system that conflicts with the actual copyright system enacted by Congress. In this way, X Corp would yank into its private domain and hold for sale information that belongs to each individual content provider — even though X Corp’s own terms and conditions says each user owns their own content.
The court was concerned that the proposed, massive regime of adhesive terms imposed by X Corp would alter the rights and privileges of the world at large or at least alter the ownership of the content that is provided by millions of X users. Such alteration would be a violation of the Copyright Act.
So, as far as the second argument — Bright Data’s selling of the data — X Corp failed because it’s preempted by the Federal Copyright Act.
There were three reasons why the state law claims (such as breach of contract) would interfere with copyright owners’ exclusive rights. First, if X’s contract were enforceable — either against Bright Data or any of X’s consumers — that would empower X Corp as a non-exclusive licensee to exclude others from displaying the user’s content. Even though they, the X users, only license their content to X Corp, the purported agreement would make that license exclusive.
Secondly, it would interfere with the exercise of fair use. Anyone can make fair use of certain copyrighted works, but X’s purported contractual terms would prevent even the discussion of whether fair use applied. And third, if such agreement were enforceable, X could be allowed to protect and keep what Congress intended to be free from restraint — and that is data itself. Data, like any other facts, don’t belong to anyone. Only the fixed expression of ideas are controlled by the Copyright Act. If X’s claims were not dismissed, X Corp could say that the personal information, the numbers, all those points of data would belong exclusively to X.
The case was dismissed, both as to the alleged breach of contract claims, and also for the allegedly improper use and access of X’s data. So, at least for now, Elon Musk does not own everything. (But take note: Big Data can keep on scraping and using any data available through X.)
