clock menu more-arrow no yes mobile

Filed under:

The dark underbelly of Amazon Prime Video

Nearly two-thirds of the videos on Amazon’s streaming service are user-generated content. Lots of them are ... odd.

A hand holding a phone with “Amazon Prime Video” on the screen. Marc Atkins/Getty Images
Rani Molla is a senior correspondent at Vox and has been focusing her reporting on the future of work. She has covered business and technology for more than a decade — often in charts — including at Bloomberg and the Wall Street Journal.

Amazon Prime Video has won lots of attention in recent years for its highly produced, critically acclaimed originals, including Fleabag, The Marvelous Mrs. Maisel, and Transparent. What’s less known is how the streaming service relies heavily on often-bizarre user-uploaded content. As the Wall Street Journal points out in a new feature, some two-thirds of the titles on Amazon Prime Video are uploaded by users, and although you might expect Amazon to be selective in what ends up on its subscription-based service, that is not the case.

We did some more sleuthing and found even more weird and potentially offensive content. It’s almost as though Amazon welcomes the bad videos, which count toward the total number of titles available on Prime Video. According to Ampere Analytics, Amazon Prime Video boasts 65,504 distinct titles — almost 10 times the 7,177 on Netflix. Users who upload videos, WSJ reports, also get a small cut of revenue based on how many people watch their videos, so there’s an incentive to upload even more. A quick glance at what turned up in a handful of search results shows that quantity can outweigh quality.

For example, feast your eyes on “Surprise Eggs Learn Colors Candy Baby Doll Bath Time,” which ostensibly could be used to teach kids colors but feels like a bad acid trip.

Then there are the conspiracy theories. Amazon has quite a few amateur-produced documentaries about what really happened on 9/11, including Dylan Avery’s Loose Change, a pseudoscientific look into how the terrorist attack was an “inside job.”

There’s also a strange two-part series called Hitler in His Own Words as well as the markedly more intriguing 3rd Reich: Hitler’s UFOs. It answers such pressing questions about Nazi aircraft like, “Was this really Nazi technology or alien technology being back-engineered by the Nazis?” (Spoiler: It was the Nazis).

And just for fun, there’s “Foot Fetish,” a video about two shoes having sex and eventually making a baby shoe. It’s described as being the Best Short Film winner in the New York Erotic Film Festival in 1973.

A short film about shoes having sex is not the worst thing on the internet. However, the trend of odd or offensive content on Amazon Prime Video stands to confuse Prime members, who pay $120 a year for access to the streaming service and other perks of Prime membership like free shipping.

We’ve come to expect off-putting content from social behemoths like Facebook and Google’s YouTube, where many regular people — and the occasional coordinated efforts from foreign governments — post their memes and videos. Amazon Prime Video, on the other hand, presents itself as a Netflix competitor, and that might lead its users to believe that the content on the platform has been vetted. To the average user, it’s not even clear that any of the content on Amazon Prime Video is user-generated, much less the majority of it. Unlike YouTube, Amazon doesn’t label user-generated content as such.

Moderation is a problem for any platform dealing with user-generated content, however. YouTube is the prime example of these ills. In recent years, the platform has attracted criticism for being rife with questionable kids’ content, some of which showed children's favorite cartoon characters in sexual or violent situations. Some of that children’s content also seems to be generated using AI to feature the keywords and topics most likely for kids to click — and for ad sellers to make money. YouTube has cracked down on creepy videos but a few still make it through the cracks.

Roku represents another unexpected source of unusual user-generated content. The company’s open platform allows any developer to create channels or screensavers for the Roku streaming platform. A couple of years ago, Vice pointed out that these easily accessed channels can also feature inappropriate content for kids, not to mention weird porn, gun violence, conspiracy theories, and doomsday cults.

But when you’re paying for a premium streaming service like Amazon Prime Video, the existence of such content is jarring. Netflix has its own conspiracy theory problem, but its users aren’t uploading their own content so it’s potentially more manageable.

Meanwhile, it’s easy to see why it might be tempting for a service like Amazon to include user-generated content. For one, it’s free, while producing or licensing content from known filmmakers is expensive. And again, inviting user-generated content onto its service also allows Amazon to boast a bigger library than its competitors and, by extension, a potentially larger draw for its paying subscribers.

Amazon does have policies forbidding offensive and illegal content. According to WSJ, the company uses both an AI tool and human reviewers to monitor content, although it’s unclear how well this process is working since we were able to find borderline content so easily. Amazon did take down several videos from conspiracy theorist Alex Jones after they were flagged in WSJ’s coverage; Jones has previously been kicked off Facebook, Apple, YouTube, and Spotify. Last spring, Amazon also pulled a bunch of videos that incorrectly linked vaccines to autism, a particularly insidious and lasting conspiracy.

In a way, Amazon Prime Video works a lot like Amazon’s marketplace, where the company has struggled with guidelines and enforcement of what is acceptable for it to carry or not. The video streaming service also suffers from a similar set of struggles that crop up when users are given free rein over what will appear on the platform. When users stand to make money by selling random stuff or simply uploading random videos, some bad stuff is bound to appear.

Amazon’s strange user-uploaded content ultimately points to an even larger challenge of content moderation. Platforms have to strike a balance between getting the benefits of free user-uploaded content with the fact that a lot of that content might not fit their standards — and those platforms might not be willing or able to invest in the tools to vet that content. As we know, nothing on the internet is ever free. Apparently, it’s also seldom normal.

Sign up for the newsletter Sign up for Vox Recommends

Get curated picks of the best Vox journalism to read, watch, and listen to every week, from our editors.