On March 16, 2020, Didier Raoult, one of the world’s leading experts on infectious diseases and the director of the medical institute Méditerranée Infection (Marseille, France), reported at a scientific conference that a combination of hydroxychloroquine and azithromycin rapidly cleans viral loads in patients with COVID-19. On the same day, the Institute published the presentation on its YouTube channel and embedded the video in its website. Over the night of March 28–29, Google’s YouTube removed this presentation! YouTube showed the message, “Video unavailable. This video contains content from Canal Plus, who has blocked it on copyright grounds.”
Simultaneously, Google and some other Big Tech platforms deleted information about hydroxychloroquine treatment from their other services. Millions of lives were lost, and there is no end yet. But this article is about DMCA 512(g), the clause that has been used by YouTube to censor information at will and on directions from third parties.
The Digital Millennium Copyright Act (DMCA) was passed in 1998. DMCA Section 512 is no less important for Big Tech than the much more famous Section 230 of the Communications Decency Act (CDA). YouTube has been using or abusing 512(g) more than other Big Tech platforms (the “Platforms”). This is how it begins:
“[Sec. 512] (g) Replacement of Removed or Disabled Material and Limitation on Other Liability.— (1) No liability for taking down generally.—
Subject to paragraph (2), a service provider shall not be liable to any person for any claim based on the service provider’s good faith disabling of access to, or removal of, material or activity claimed to be infringing or based on facts or circumstances from which infringing activity is apparent, regardless of whether the material or activity is ultimately determined to be infringing.”
Paragraph (2) imposes very mild requirements. The breadth of the 512(g) clause is breathtaking,
Today, one does not need to copy someone’s video or text to be reasonably accused of a copyright violation. For example, a video might be infringing somebody’s copyright, even by simply showing a purportedly copyrighted object (e.g., a painting, architecture, or a live performance) in the background.
However, 512(g) protects the service provider even if the third-party infringement claim was bogus and even if there was no third-party claim (“infringing activity is apparent”). There is a formal requirement of good faith, but the absence of good faith is hard to prove. In addition, 512(f) limits the liability of the service provider even for outright misrepresentations in banning the content to actual damages and attorney fees incurred by wronged content poster. The service provider has no liability to readers or viewers under Section 512.
Further, Platforms’ protections under DMCA 512(g) and CDA 230(c)(2) overlap. A platform facing a lawsuit can assert either or both paragraphs, and to claim a mistake. If the plaintiff gets through this gauntlet, s/he must prove damages. That would take years, and the Platform can deplatform, ban, and organize boycotts against the plaintiff as much as it wants. The worst case for the Platform is paying the damage and attorney fees of the victim. In other words, censorship is free.
Thus, 512(g) became an unconstitutional censorship privilege for a few Big Tech corporations.
How did we get here?
The DMCA was passed in 1998. Most of the service providers existing then were Internet access service providers (ISPs) and web hosting providers. The following was true for these services at that time, and probably was assumed for other service providers, relying on 512(g):
- The service provider was interested in providing the best experience to each consumer. In other words, the provider was on the side of the consumer.
- Service providers were in a very competitive market. A consumer who was unhappy with the provider’s DMCA policies had a broad choice of alternatives.
- The government did not attempt to influence the providers to censor their users under protection of 512(g).
- Service providers were not connected to government, not influenced by government, not dependent on government, not selected by government, and not closely related to any political party.
- Providers did not have excessive power; providers’ power was balanced and checked by organizations representing their consumers. Thus, it was possible for consumers to find out and prove in court the bad faith of a provider.
Only under all these conditions was 512(g) constitutional. Possibly, another important factor was the small part that the Internet played in the public debate in 1998.
Today, none of these conditions is true for the Big Tech platforms (including Google, Facebook, Twitter, Microsoft, and Apple):
- Big Tech companies have hundreds of millions, or even billions, of users around the world. Any one consumer is lost in this ocean. Also, Big Tech is openly contemptuous of a quarter of its consumers, the “deplorables,” and Big Tech is more interested in satisfying the DNC, CCP, the European Commission etc. than its consumers. Despite this, the Obamanet has forced all US Internet users to subsidize Big Tech through ISP fees, whether they use their services or not.
- Each Big Tech company is a monopoly (or, worse, a closed network) in the main services it provides. Also, they collude among themselves on the censorship.
- The Biden-Harris administration openly pressures Platforms to censor their consumers and third parties under any pretext.
- Big Tech is the partner of the US federal and state governments, many foreign governments, and the Democratic party. The Big Tech corporations have been hand-picked by the Obama government, and received multiple benefits from it, including exclusive opening of government accounts with the Platforms and Obamanet. After a covert period, the Platforms openly partner with the government in “fighting misinformation,” “Russian propaganda,” “securing elections,” violating the US equivalent of the Nuremberg code This qualifies Big Tech as state actors, even in the absence of 512(g). However, even if we close our eyes and assume that the Platforms are not state actors without 512(g), they become state actors when government influence or connections are combined with 512(g).
- Big Tech has enormous power, and even big states like Texas and Florida have difficulty doing anything about them. Individual consumers are powerless.
The banned video
At the time of the ban, the video had 1.4 million views. It was also on the channel of a large and well-known medical institute. This ban cannot be explained away as a mistake.
Approximately the next day (by March 30), Google changed the explanation to “An error occurred. Please try again later.” This stresses how Big Tech abuses the situation when it has no obligation of disclosure.
Also, notice the “little lie” in the March 29 disclosure: YouTube claimed that Canal Plus (a French TV channel) blocked the video, although only YouTube could block on YouTube. Canal Plus could only send an infringement notification. It is very unlikely that Canal Plus had sent one, especially because nothing in the video might be claimed by Canal Plus as copyrighted (the video seems unchanged before and after the ban). It is likely that Google employees made a false record in a database to block information about this COVID-19 treatment because it had been touted by President Trump.
The video was back by April 7, but by this point, it was too late.
 Paragraph (2) has no effect whatsoever on bans made by YouTube or another Platform without an infringement notification. Even if the Platform intends to allege receiving an infringement notification, it can keep the ban for at least 10 days – an eternity in the current times. Finally, the Platform’s obligations are expressed in phrases like “takes reasonable steps promptly to notify the subscriber”.
 In re Cincinnati Radiation Litigation, 874 F. Supp. 796 – Dist. Court, SD Ohio 1995