Skip to main content
  • State of the Internet

Why the DNS should be excluded from Canada’s online safety framework

By Georgia Evans

In March, the Government of Canada announced a new expert advisory group as the next step in their process to develop legislation to address harmful content online. The Online Safety Expert Panel was tasked with providing advice on a legislative and regulatory framework to best address harmful content online. The panel completed a series of workshops, with guiding worksheets developed by the Department of Canadian Heritage—the line department responsible for the creation of this legislation.

Without a process for the panel to complete external stakeholder engagement, CIRA monitored the panel’s activities and submitted specific recommendations to Department of Canadian Heritage officials overseeing the file based on the outcomes of the advisory group’s deliberations.

CIRA is uniquely positioned to provide recommendations regarding online safety for a number of reasons. CIRA is the sole operator of the .CA registry, its associated .CA domain name system (DNS) and is a cybersecurity provider. CIRA is also an industry expert on internet architecture and involved in global fora for multi-stakeholder internet governance offers.

Central to our submission is the fundamental belief that any legislation intended to advance online safety in Canada must do so without threatening the open, non-proprietary standards on which the internet was built, and without compromising the technical infrastructure on which it operates. We also argued the following:

  1. Excluding the DNS from the scope of internet regulation is the correct approach

In the previous 2021 Technical Paper for online harms, and throughout the panel’s worksheets, intermediaries who provide domain name services were to be excluded from the regulation of illegal content. Excluding the DNS from internet content regulation is the right approach.

Intermediaries providing domain name services operate at the lower layers of the internet stack—the network and transport layers—and as such, do not transmit or store the content of websites, nor do they have any influence over the content of websites. Any DNS-level action to mitigate and remove the content of websites would involve removing or preventing an entire domain name from being resolved, which would be a disproportionate response to managing the implicated speech.

  1. Illegal content is best managed at the upper layers of the technology stack

Unlike DNS providers, intermediaries who operate at the application layer—like social media companies and web hosting companies—are much nearer to where content is posted and have a closer relationship to the people who post it. As a result, they can intervene when necessary to remove specific illegal content with precision.

For example, if a user was to post hate speech or terrorist content on Facebook, then the company’s internal content moderators would be able to remove the content from the platform. In CIRA’s view, this is the most precise and accurate way to make illegal content inaccessible to Canadians, as opposed to DNS-level action, which can’t make specific content inaccessible without disrupting access to all of Facebook.

  1. Excluding the DNS from the online safety framework aligns with international best practices

We also noted in our submission that excluding the DNS from the online safety framework aligns with the approach that the European Union (EU) and United Kingdom (UK) have taken for internet content and services regulation.

The EU’s Digital Services Act (DSA), for example, distinguishes between categories of intermediaries and imposes obligations for different types of services providers that are proportionate to the size, impact and risk posed by their services. Under the DSA, the largest responsibility for curbing illegal content rests with very large online platforms and very large online search engines that reach over 45 million users in the EU, as they pose the most risk in the content’s dissemination.[1] These are just some ideas that deserve consideration as Canada’s approach to online safety takes shape.

  1. Website blocking by ISPs is a tool of absolute last resort

In the 2021 Technical Paper, the Government had proposed that a Digital Safety Commissioner should have the authority to apply to the Federal Court for an order for telecommunications service providers to block access “in whole or in part” to an “online communications service” – i.e. Facebook, Twitter, or a website hosting impugned content, if the provider of the service repeatedly did not comply with orders to remove child sexual exploitation content or terrorist content.[2]

While the panel workshops did not discuss the issue of website-blocking, CIRA urged the Government to ensure that court-ordered website-blocking only be included as a tool of absolute last resort for addressing illegal content online. This, especially when, there are more proportionate responses and intermediaries available to remove the content with precision.

At CIRA, it is our firm belief that the internet is central to Canadian life and vital for the full participation of all Canadians in our economy, society and democracy. That’s why building a trusted internet for Canadians is a key part of our mission. We commend the Government of Canada for its commitment to fostering a safe online environment for Canadians and will continue to support its ongoing efforts to combat illegal activity online while protecting Canadians’ freedom of expression.

[1] European Commission, “The Digital Services Act: ensuring a safe and accountable online environment,” European Commission,  

[2] Canada, “Technical Paper,” Canadian Heritage, 2021,

About the author
Georgia Evans

Georgia is a Policy & Advocacy Analyst at CIRA and is very passionate about internet governance and digital policy.