A few weeks ago, I posted a video about how to bypass paywalls on YouTube: https://youtu.be/qQ6nn9pSiiY.
But a week later, it was taken down:
It would seem that under the “Fair Use” law, that video should’ve been allowed to stay up. However on YouTube, (unlike regular law in the US), you’re guilty until proven innocent, meaning that if someone claims your video is in violation of copyright law, then your video gets removed right away, even if the law would consider your video to be “Fair Use”.
Given the presence of lawyers in my immediate family and circle of friends, it would seem reasonable for me to gather information about the applicable law, and make a case for why my video about bypassing paywalls should be allowed to remain on YouTube.
Here’s a good summary of “Fair Use”:
Let’s go through the four factors of “Fair Use” consideration, and see if my video applies:
- “The purpose and character of the use”. My video was about teaching various concepts, such as how to disable JavaScript in the web browser to prevent a pop-up from appearing when scrolling down on a web page. I also taught that https://archive.org/ (The “Wayback Machine”) is a site that makes backup copies of millions of sites on the internet. The web page in question, (where I demonstrated bypassing a paywall), had already been archived by The Wayback Machine. (Original: https://www.alltrails.com/trail/us/vermont/thundering-brook-falls. Archived copy: https://web.archive.org/web/20221101005323/https://www.alltrails.com/trail/us/vermont/thundering-brook-falls.) In the video I also taught about web-scraping. This would be considered a “transformative” situation, as I used an automation framework that I created, https://github.com/seleniumbase/SeleniumBase, in order to bypass the AllTrails’ bot-detection systems (provided by https://datadome.co/). In an earlier ruling, which set a legal precedent (https://brightdata.com/blog/web-data/court-rules-in-favor-of-bright-data-in-meta-v-bright-data-case) it was determined that web-scraping public data is legal as long as you don’t have to log in to a site to collect the data. In my situation, all the data that I demonstrated web-scraping on was publicly available without needing to log in. Also, my video was for non-commercial use. These points would appear to satisfy Factor 1 of Fair Use.
- “The nature of the copyrighted work”. The original webpage was both factual and already published to the rest of the internet. This would appear to satisfy Factor 2 of Fair Use.
- “The amount and substantiality of the portion used in relation to the copyrighted work as a whole”. A single webpage of AllTrails was used to demonstrate the effect of disabling JavaScript, as well as performing web-scraping. Only a small number of the over 1,000 reviews of that trail (Thundering Brook Falls) were actually visible on that webpage while being viewed/scraped. And Thundering Brook Falls is only one trail of the thousands of available trails that exist in the US. These points would appear to satisfy Factor 3 of Fair Use.
- “The effect of the use upon the potential market for, or value of, the copyrighted work”. My target audience is people who do web-scraping, as opposed to the target audience of AllTrails, which is hikers. The principles that I taught in my video are meant to be applied to any website with a paywall that appears after performing an action, such as scrolling on a page. This is not specific to AllTrails, although AllTrails is a excellent example of a site that tries to block content via JavaScript. The criticism mentioned of having JavaScript-as-a-security-measure are done in a transformative way. As for market impact, this might actually get more people to use the AllTrails website, because many people would leave right away if they had to sign up for a service to reach the information. Also, discussion on YouTube about bypassing paywalls is not a new concept. (Eg. https://www.youtube.com/watch?v=HmnA0wjp6j4, “Bypass website paywall banners to view the content for free”, is a video by someone else with over 70,000 views on YouTube, which demonstrated the concept of disabling JavaScript to bypass paywalls on The New York Times website, which was allowed to remain on YouTube.) This feature of disabling JavaScript has been available in Chrome browsers for a very long time. These points would appear to satisfy Factor 4 of Fair Use.
As for past precedent, since https://www.youtube.com/watch?v=HmnA0wjp6j4 is allowed by YouTube (discussed above), then my video should be allowed too. The main difference there is bypassing paywalls on The New York Times website vs. bypassing paywalls on the AllTrails website. (The same trick with disabling JavaScript from Chrome Settings was used for both).
With regards to future precedent, there’s the potential slippery slope if saying that you can’t record video at all during a screen-share (and post it to YouTube) if you’re visiting a website that you don’t own. (Unless that’s already covered by “Fair Use”, which I already discussed above.)
If there was an issue with the web-scraping tutorial from my video, then we should look to the legal ruling from https://brightdata.com/blog/web-data/court-rules-in-favor-of-bright-data-in-meta-v-bright-data-case, which established that web-scraping is legal in the eyes of the courts as long as you’re only scraping public data while not logged into a website.
With regards to The Internet Archive (The Wayback Machine), the particular page of AllTrails that I used for my example was already part of The Internet Archive: https://web.archive.org/web/20221101005323/https://www.alltrails.com/trail/us/vermont/thundering-brook-falls.
Given that, and the important details of this legal ruling: https://www.pinsentmasons.com/out-law/news/use-of-web-archive-did-not-constitute-hacking-says-us-court – “Use of web archive did not constitute hacking, says US court”, this directly applies to my situation because the ruling in that case stated:
“The use of web archive The Wayback Machine did not constitute hacking in the case of a law firm which used the web archive to see pages which owners did not want it to see, a US court has ruled.”
Also related to the Internet Archive: Internet Archive Helps Secure Exemption To The Digital Millennium Copyright Act.
More formally, Internet Archive has successfully advocated for an exemption to the Digital Millennium Copyright Act (“DMCA”).
With regards to the email notification I received from YouTube about the video removal, there’s an error (misrepresentation) in AllTrail’s submission for the removal request on my video:
I’m not sure how AllTrails can claim that my “Entire video” is all content belonging to them. Some of my video included home video clips that I made, selfies that I took, videos of me navigating through the SeleniumBase GitHub page, and various photos of me with family, such as a photo of me and my wife in Disney World. That “Entire video” designation can be a serious problem with regards to YouTube’s Content ID system, because then my personal content can get labeled as not mine. (There is an option to select parts of a video in that step, but AllTrails went ahead and selected “Entire video” instead.) The legal term for that is “misrepresentation”.
YouTube’s email to me also mentioned that: “If you believe you’re not at fault in one or more of the instances above, you can appeal this removal by submitting a counter notification.” But according to the Reddit conversations of several people, that hasn’t been nearly as effective as going vocal about it. Hence the reason for this blog post.
For anyone interested in a legal discussion about this, I created a room in the SeleniumBase Discord Server: https://discord.gg/Tr3vnqx8xG. (Update: Looks like there’s over 800 members of SeleniumBase Discord now!)
If you’re from AllTrails, I’d certainly like to hear what you think. If you think that my video doesn’t qualify for “Fair Use”, despite the points I’ve mentioned above, then I’d like to hear why. Or, if you agree that my “Fair Use” points carry validity, then I’d love to see my video back up on YouTube. If you still don’t believe it qualifies for “Fair Use” based on the points I made above, then I’d like to hear your opinion on why you think the situation in my video is different from https://www.youtube.com/watch?v=HmnA0wjp6j4, which was deemed “Fair Use” (where someone demonstrated bypassing a paywall on The New York Times website while displaying content from the page to prove that the paywall was bypassed successfully).
Since https://seleniumbase.com/ doesn’t fall under YouTube’s jurisdiction, my video is coming back up here so that people can decide for themselves about whether or not the video qualifies as “Fair Use”:
Based on the rules of the “Fair Use” legal doctrine, it appears that the content in my video qualifies as “Fair Use”.
Special thanks to family/friends who provided me with free legal consultation to help me compose this blog post.