This is the first in a series of regular posts where I’ll dive into an interesting opinion to understand which facts and sources of evidence had the most impact on the judge’s ruling.
First up, it’s the recent Parler v. AWS matter, where Parler sued Amazon for refusing to continue to host the right-wing social media platform on AWS, after Parler users made thousands of shockingly violent posts after the January 6 Capitol insurrection. The Seattle Times had a nice summary of the lawsuit and the hearing that preceded the order we’re about to unpack. But while the press focuses on the first amendment-y aspects of the case, Parler’s loss came down to the fact that they relied on dramatic allegations instead of cold, hard facts.
Parler sued Amazon seeking a preliminary injunction that would force AWS to put Parler back online, ASAP. (For the non-lawyers in the house: a preliminary injunction means that Parler needed to prove to the court that it was decently likely to win on the merits once the case progressed further, so that the court should basically just give them what they want now without having to wait for the end of the lawsuit.)
First Up: The Sherman Act (Antitrust) claim:
Parler tried to argue the antitrust equivalent of “but mom, you let him do it!” Parler was claiming that Amazon allowed Twitter to get away with hosting similar violent content, and that Amazon’s actions with Parler were part of a nefarious plot to help Twitter succeed.
Evidence faceoff:
• Parler didn’t submit any factual evidence that Amazon and Twitter conspired to get Parler out of the way in order to reduce Twitter’s competition. Womp womp.
• Meanwhile, Amazon submitted a sworn declaration by a high ranking Amazon executive who explicitly denied that Amazon and Twitter ever had a single conversation about Parler, and noted that to do so would violate Amazon’s policies of never sharing one customer’s information with another. He also emphasized his powerful executive role and said he never authorized anyone on his team to have this kind of conversation, so therefore it couldn’t have happened.
• Also, Amazon dunked on Parler with its evidence showing that technically, it didn’t even host Twitter yet. An executive stated in a declaration that the agreement to work with Twitter was signed in Dec 2020, but Twitter’s feed wasn’t yet on AWS. Amazon used this fact to argue that it didn’t even make sense to compare their relationship with Twitter.
Takeaways: Amazon had to prove a negative here–that they never discussed something internally. That’s a tough thing to prove when extensive discovery hasn’t yet occurred. They decided the quickest way to “prove” this was to have an exec explain the internal policy against discussing customer data, and swear that since he didn’t personally have or authorize anyone else to have convos with Twitter, it must not have happened. Is this airtight evidence? Nope. But since Parler submitted nothing, the court held it was good enough.
The argument about Twitter’s feed not yet being on AWS is a great example of using a technical piece of evidence to help the judge zoom out and see the practical realities of Amazon’s position. Amazon is explaining that while it has to take responsibility for content it hosts on its platform, Twitter isn’t up there just yet. At this particular moment in time, Parler is a (rotten) apple, while Twitter is an orange. Interestingly, having made this argument, once Twitter’s feed is live on AWS, Amazon better be able to show evidence that it’s treating Twitter the same re: shutting them down for failure to moderate violent content.
Next up: The Contract claim:
Evidence Faceoff:
• Amazon and Parler each claim that the other company violated the terms of the AWS contract. The judge’s ruling here came down to one simple fact–that Parler simply didn’t deny Amazon’s argument that Parler’s content violated the agreement.
• Parler tried to focus on a contract technicality, that Amazon needed to notify them and give them 30 days to fix any issues with the violent hate speech.
• But the judge wasn’t hearing it because Parler didn’t even bother to argue or submit evidence showing that the violent content wasn’t in violation.
Takeaway: Again, Parler whiffed it by simply not providing any evidence to contradict Amazon’s argument about the content violating the contract. Amazon’s strategy was simple and powerful–show the court examples of the horrific, violent content Parler allowed on its platform. How could Parler have engaged more with this evidence? Perhaps they could have at least tried to analyze for the court a selection of the troubling posts, highlight any gray areas in whether they should be removed, and give a specific discussion of the timeline of how and when such posts would have been taken down. And as a side note, I wonder whether their counsel didn’t want to be seen as supporting violent extremists by making this argument and implicitly debating the merits of these violent posts.
It all added up to a lack of facts in the record that would support a court forcing Amazon to host this violent content. This opinion is a pretty classNameic example of why litigants need to give the court something substantive to back up their lofty arguments, like documents or declarations with sworn testimony. This is especially important when seeking a preliminary injunction (which has a seriously high bar before a judge will grant one–the case hasn’t run the usual course yet).
And the case also shows that before discovery, when the parties can’t supply the court with lots of emails and texts they found to support their claims, a tightly worded declaration by a company executive can be pretty powerful evidence.
For more litigation nerd content, and to see a delightful selection of lawyerly memes, you’re invited to join our private community group at Let's Be Brief.
