Over the past couple of years I have written, as have others, about the abuses generated from the free-ride that large internet platforms have been given as a result of their exemption from liability for abusive or sometimes illegal content carried on their services and disseminated by them. They have thrived in this permissive environment. In the US, the cornerstone of this immunity is Section 230 of the Communications Decency Act, 1996. This piece of legislation, passed at the dawn of the internet era, has come under increasing scrutiny in the US–most recently because of the antics of Donald Trump.
Trump’s online comments have long been suspect with regard to accuracy, but it was his comments justifying violent, seditious activity such as the attack on the Capitol Building on January 6, that finally forced the major platforms to act. Facebook/Instagram, Twitter, Youtube, Reddit, Snapchat and others all took action to block his access to their services, either temporarily or on a longer-term basis. The flow of alternate facts and incitement to violence was finally turned off. It has taken far too long–and needed an incident that shook US democracy to its core–for the major social media platforms to finally “man up” and exert some control over the harmful, even dangerous, content propagated by Trump. While on this occasion they acted, on too many others the platforms have hidden behind the immunities provided by Section 230. This legislation is 25 years old this year; it is high time for a serious update.
Why is Section 230 a Concern?
Why am I writing about this in a copyright blog when intellectual property infringement is specifically carved out of Section 230 immunities for internet platforms? There are other laws in the US (viz. Section 512 of the Digital Millennium Copyright Act, known as the DMCA) that provide safe harbours to the platforms when copyright infringement takes place, provided they take certain actions. (Section 512 is also under review because in the view of many it has not operated as intended to protect copyrighted content online). Section 230 is relevant because of the broader need for internet intermediaries (platforms) to assume responsibility for the content they propagate and profit from, especially when they are well aware of the abusive content they are enabling. That need for greater accountability and responsibility extends into many areas, including respect for copyright and creators.
The lack of responsibility—enabled by Section 230—that has been a feature of the industry for many years is now catching up to it. The immunity should never have been so broad or widespread. Section 230 has been used as the shield to protect websites promoting child sexual exploitation, selling weapons to individuals prohibited from purchasing them in the offline world, and promoting hate-speech; it has even allowed businesses like Airbnb to ignore municipal regulations on home rentals. Through its broad immunity shield, it has indirectly enabled revenge-porn, doxing, cyber-bullying and defamation. No wonder there are calls for review.
Not only is the new US Congress likely to scrutinize the protections that internet platforms have enjoyed—and often abused—for the past 25 years, but the unfettered right of these platforms to host defamatory or inappropriate content is being challenged in other forums as well, such as courts outside the US. On several occasions, national courts have sought and exerted jurisdiction over cases of abusive online behaviour tolerated by the platforms that has affected their nationals.
Twitter Challenged in a Canadian Court
The most recent example is the lawsuit brought against Twitter in the British Columbia (BC) Supreme Court by Frank Guistra, among other things the founder of Lionsgate Entertainment. He has accused Twitter of publishing and then failing to remove false and defamatory tweets related to him. The tweets in question accused him of being involved in “Pizzagate”, the spurious conspiracy-theory about pedophilia that arose during Hilary Clinton’s campaign against Donald Trump in 2016.
Twitter had argued that the BC Court did not have jurisdiction, and wanted the case heard in California, where Guistra has a residence and where Twitter maintains its corporate headquarters. Guistra, who is a Canadian citizen and whose career in the financial, mining and entertainment industries has been primarily pursued in Canada, maintains his home in Vancouver. Twitter wanted the case heard in California for obvious reasons—they could hide behind Section 230, as well as argue the First Amendment to the US Constitution. The First Amendment provides for freedom of speech and freedom of the press, amongst other things, although it applies only to the role of government. “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press…”. It doesn’t stop Twitter from moderating content on its platform.
There are several interesting aspects to this case; that of jurisdiction, the geographic reach of a potential ruling on the case, and probably most important, whether a social media platform like Twitter can be held liable in Canada for defamation arising from content published on its service by third parties, its users. Related to this last point is the question of whether Section 230 has extended its long arm into Canada as a result of Article 19.17 of the USMCA/CUSMA, the updated version of NAFTA that entered into force in July 2020. Noted copyright lawyer and blogger Barry Sookman has an extensive post on the various legal ramifications of this case, which is well worth reading.
Does Canada have Jurisdiction?
I won’t repeat the well-documented legal arguments in Sookman’s blog posting but want to touch briefly on the jurisdiction question and then take a deeper dive on Section 230. The court asserted territorial competence because the harm to the plaintiff, a BC resident, occurred in BC (as well as elsewhere). It then determined it had jurisdiction–despite the existence of US law on the subject–because under US law the plaintiff would have had no cause of action (because of Section 230) for the harms suffered in BC. With the venue issue settled (unless there is a successful appeal), the really interesting point is whether Twitter is liable in Canada for defamation.
Can Twitter Rely on USMCA/CUSMA Article 19.17
Sookman points out that, in a Canadian context, strictly passive intermediaries (bulletin boards, social media platforms, etc.) will not be held liable for publishing defamatory content posted by users unless they know, or should have known, that their services are being used to distribute such content and they take no steps to remove it. Guistra had specifically requested that Twitter remove the defamatory tweets, and block future ones. Allegedly it did not do so fully, despite repeated requests.
Assuming no appeal, or an unsuccessful appeal, the case will be heard in Canada. This brings Section 230 back into the picture, specifically the language in USMCA/CUSMA Article 19.17, which reads in part;
“…no Party shall adopt or maintain measures that treat a supplier or user of an interactive computer service as an information content provider in determining liability for harms related to information stored, processed, transmitted, distributed, or made available by the service, except to the extent the supplier or user has, in whole or in part, created, or developed the information.”
This language almost didn’t make it into CUSMA because of last minute objections by Nancy Pelosi and House Democrats at the time the Trump Administration was seeking Congressional approval of the negotiated USMCA deal. Other changes were made to the treaty at the request of the Democrats, changes that Mexico and Canada subsequently agreed to, but Pelosi’s eleventh hour opposition to inclusion of Article 19.17 was too little, too late, and the language, which mirrors parts of Section 230 but with modifications, stayed in. It is the sort of measure that should not be in a trade agreement, and many in the US are opposed to exporting it to other countries.
There is a qualifier (footnote 7) to 19.17 which reads;
In other words, Canada was not obliged to pass any new legislation as a result of the USMCA/CUSMA commitment and can comply through the application of existing law.
Will the USMCA/CUSMA language shield Twitter from Giustra’s suit? It will be interesting to find out. The question of the extent to which Article 19.17 will bear on defamation cases in Canada involving internet platforms is an open one, as I wrote just over two years ago (“Did Canada Get Section 230 Shoved Down its Throat in the USMCA?”). You can read that blog for yourself, but my conclusion was that;
“…while the Parties have agreed under the USMCA to not treat a platform as the creator of content, in other words as a primary publisher, platforms are still liable under the Canadian common law as secondary publishers when they knowingly publish the contents of a primary publisher that is, for example, defamatory.”
“Far from having Section 230 type safe harbours shoved down its throat, Canada protected its ability to regulate platforms and protected the ability of the courts to take action against platform abuse where and when required.”
Does Article 19.17 Cover Secondary Liability?
This interpretation is based on the explicit wording of Article 19.17. While the article prevents the Parties to the USMCA from treating an “an interactive computer service” (like Twitter) as an “information content provider” (publisher) for purposes of liability for content “stored, processed, transmitted, distributed, or made available” on its service (unless it has created the information itself), that protection is related solely to not treating the interactive computer service as a content provider. This expressly does not exclude secondary publishers from potential liability. Rather, they may be liable if they have published content from someone else, have knowledge of what they have published, and have refused to remove it. That is why the common law clarification in the footnote is important. It confirms that liability as a secondary publisher is not being changed.
Article 19.17 and Section 230: Some Significant Differences
The precise language of Article 19.17 is also interesting because, unlike Section 230, it does not limit access to injunctive relief. It does not exclude the issuance of injunctions, or court-ordered equitable relief whereas Section 230 has been interpreted to prevent the issuance of injunctions against internet intermediaries.
Clarifying Section 230’s Reach into Canada
The Giustra case should help clarify the extent to which Section 230 has reached into Canada through the back-door of Article 19.17. The outcome will be important. Meanwhile, Section 230 is unravelling in the US, and there are moves to amend or get rid of it, notwithstanding Article 19.17 which of course applies as much to the US as it does to Canada and Mexico (although Mexico negotiated a three-year delay in implementing it). If the US Congress intends to forge ahead and amend or repeal Section 230 in order to start holding internet platforms accountable for the harms that they disseminate through unmoderated content posted by users, regardless of the USMCA, then one wonders how much credence courts in Canada will or should give to the trade agreement’s “obligation”, particularly since no legislation was needed to implement it. As I noted above, Canada was free to comply with the commitments in Article 19.17 through the application of existing legal doctrines (i.e. common law principles) as applied through judicial decisions.
Does Article 19.17 have any Legal Effect?
Vivek Krishnamurthy and Jessica Fjeld, two legal scholars who published a research paper on issues related to the implementation of Article 19.17 in the US and Canada have pointed out that;
“The Supreme Court of Canada has found that international treaties are of no force and effect beyond what is provided in their implementing legislation. Since there is nothing in Canada’s USMCA implementing legislation that refers directly to Article 19.17, the provision appears to have no direct domestic legal effect. However, the Supreme Court has also held that it will consider Canada’s international obligations and customary international law as interpretive aids when the meaning and effect of a provision is called into question.”
However, it has since been pointed out to me that the above interpretation is not uncontested in legal circles, and is at variance with the conclusion from the 2007 Regina v Hape (SCC) case that there is a presumption that Canada’s laws will confirm to treaties;
...courts will strive to avoid constructions of domestic law pursuant to which the state would be in violation of its international obligations, unless the wording of the statute clearly compels that result… (R v Hape,  2 SCR 292 BOA Tab 92 at para 53).
Section 230 Reform in the US
We will have to wait to see what happens with the Giustra case with regard to the applicability of a Section 230-like safe harbour in Canada. In the meantime, calls for Section 230 reform continue to get plenty of traction in the US, such as this thoughtful analysis published by Dr. Mary Anne Franks, Professor of Law at the University of Miami, and concerns about under-filtering identified by University of Virginia School of Law professor Danielle K. Citron. What happens in the US may eventually influence what happens in Canada with regard to Article 19.17 because if it becomes a dead letter in the US, one can hardly expect that it will be rigorously applied in Canada (or Mexico).
© Hugh Stephens, 2021. All Rights Reserved.
This blog has been updated to include reference to the Regina v Hape case with respect to the impact of international treaties on domestic law in Canada.
Further update. In December 2021 the BC Court of Appeal rejected Twitter’s appeal challenging BC jurisdiction. The trial will be held in British Columbia.