The Law Commission of Ontario (LCO) is in the midst of a project to reform defamation law in light of the internet age. As part of its work, we were commissioned to write a paper with recommendations for reform on the topic of intermediary liability and the meaning of publication in defamation law.
In this post, we offer a brief summary of our findings and recommendations.
Given the breadth of the meaning of publication in defamation law, internet intermediaries such as internet service providers, website hosts, search engines and social media companies may be liable for defamatory content posted by third parties. In Canada, there is no relevant legislation, so the question is one of the publication element at common law.
A review of the case law in various jurisdictions suggests at least two problems. First, the doctrine sometimes results in defendants being found to be publishers where many would not think their conduct sufficiently blameworthy to ground liability. Second, the law is complex and confusing. Courts sometimes conflate different lines of doctrine – for example regarding secondary publishing and publishing by omission. The result is an emerging case law that is sometimes unprincipled and, in our view, is ill-suited to dealing with the issue of internet intermediary liability in defamation.
We recommended to the LCO that the publication element of defamation be redefined so as to require a deliberate act of communicating specific words. The common law distinction between primary and secondary publishers – that is, between publishers who have knowledge and control over content and those who do not – would effectively be abolished, since the latter would no longer be publishers at all. Similarly, the doctrine of publication by omission, in which one is a publisher for failing to remove content posted by others, would be abolished. Only those who deliberately communicate with knowledge of specific words at the time of publication should be treated as publishers. This change would not be limited to the internet intermediary context but rather to the tort of defamation as a whole.
That said, intermediaries incentivize and often profit from content and have considerable power to mediate between those who post content and those who object to it. We therefore recommend that intermediaries be required to follow certain procedures for handling defamation complaints. It is our view such procedures should be incentivized and harmonized through defamation legislation.
In examining intermediary liability models in other jurisdictions, the pendulum swings between broad immunities models (USA), stricter safe harbour models (better known as notice and takedown) (evident in Europe), human rights frameworks (evident in Brazil and civil society frameworks such as the Manila Principles), and Canada’s middle-path notice-and-notice regime for copyright law. With the notice-and-notice framework an intermediary does not risk liability for copyright infringement, but rather risks imposition of statutory damages for failure to pass on notices of copyright infringement to the relevant user.
We recommended to the LCO a bespoke framework that we call “notice-and-notice-plus”, modeled on copyright’s notice-and-notice regime. We drew from Canada’s copyright framework, because the underlying rationale of its intermediary liability provisions aligns with the conclusions in this report, namely that intermediaries should not be liable for the unlawful acts of third parties. However, given the significant harm to reputation caused by continued circulation of defamatory content online, we recommend that there be mechanisms for removing allegedly defamatory content in narrow circumstances. Drawing from the use of statutory damages in the notice-and-notice system under Canada’s Copyright Act, the risk to the intermediary for failure to comply with the rules should be a fine.
Our recommendations apply to intermediaries that host content (e.g. social media companies, blog hosts, owners of sites allowing third party comments). While the paper provides more detail on the proposed framework, such as issues of incentivizing corporate responsibility, discretion, good faith, provision of reasons and anonymity, to name a few, the basic rule is the following. Upon receiving notice that they host allegedly defamatory content, intermediaries would have to forward the notice to the third party that created the content complained of. Only if the third party fails to respond within a reasonable time should the intermediary remove content. If the third party responds and claims the communication is lawful, the plaintiff would have to get a court order for content removal.
Our suggested framework is grounded in three principles: applicable rules should be human-rights based; should enable innovation; and should serve to encourage corporate social responsibility. For example, removing the potential for intermediary liability decreases the likelihood that intermediaries will remove potentially lawful content simply to mitigate their risk. This promotes freedom of expression. Corporate responsibility is encouraged by taking into account terms of service and other management of its services when assessing suitability or amount of a fine.
We invite readers to read the full paper and provide feedback either to us or the LCO.