Canada’s COVID Alert App Needs to be Shut Down. Here’s Why.
Tech Experimentation, Political Consequences, Technosolutionist Policy
Over the course of the weekend, a fellow tech community member shared the unfortunate and common news that his family had contracted Covid. He was asking if there was anywhere to report their cases, and in doing so mentioned that Canada’s COVID Alert app required PCR test results to log it there. This flipped a switch for me because I knew that for the past few months, the general public couldn’t access PCR tests, they weren’t available because of capacity issues. Then the next domino fell: if the app required PCR tests, and people couldn’t get PCR tests, how was the app even still up and running? PCR testing capacity isn’t a small issue or insignificant requirement of the app, it’s the requirement that the app’s core function is built around.
So It’s Floating Around Out There — What’s the Big Deal?
I can hear some of you thinking, who cares. This was a test of an idea, it didn’t work, so what if it’s hanging around? Two main reasons:
The first reason it matters relates to the future of government tech interventions and their surrounding politics — it’s much less about the app directly. The issue for future is what happens if we don’t hold the government accountable for this experiment today. And this was very much an experiment. We have to force the government through this wind-down process because they can’t be allowed to do this kind of wide-scale experimentation lightly. Having to publicly work through the wind-down of this app is a set of real political consequences for the sitting government.
These consequences need to be remembered both by them and by us for the next time this kind of a tech approach to anything gets put on the table. And it will. These consequences also need to be understood in terms of what, and I cringe as I type this (see asterix below), responsible innovation looks like. If the government wants its capacity and intent to support good public technology to be taken seriously, they have to commit to managing the entire life-cycle of their services and applications.
*I am writing this post without devolving into how the policy considerations for this app were not well-considered when it was launched, and how the successful technical execution of code is a very different thing than successful consideration of technology in practice. This app had myriad problems from the start, but that is not the point of the post. But it’s also near impossible for me to put the terms “good public tech” and “this app” together in the same sentence because they do not belong together.* moving on.
The second reason it matters is because if these kinds of apps don’t get removed they can become normalized into persistent infrastructure. Their purpose can be changed to suit future needs. There is a path dependency and a psychological desire to make some use of this investment. In other words, there are all kinds of unintended consequences to leaving something like this lying around, and no shortage of people incentivized to invent them.
There is a pathology in technology to find some kind of “public good” in every situation. The current status of this application being in a good number of people’s phones across the country, with a tie-in to Google and Apple through the operating system, should be hustling us to get this app uninstalled before it morphs into something else that is pitched with a moral frame of “protecting each other” again.
The noise around this app was loud in the beginning. It was high-profile, it included celebrities and communications strategies. Now the challenge that we face is one I’m finding is perfectly consistent across tech interventions: time and accountability. This is the stage at which public intervention has evidence on its side. This is the stage at which all these grand and moral “what-ifs” cease to exist. This is the moment in time where the story needs to be memorialized in the government’s practice, and in our voice. And finally, this part of the life-cycle, dealing with the app’s maintenance, putting something away neatly, tidily, and with clear communication, is what any government has to show it can do in order to be taken seriously in future.
Putting the App Away — Conflicting Stories and Intentions
The process of winding this app down, often called sunsetting, was promised as something the federal government was committed to thinking about, in words only apparently, when it launched the app. This was always supposed to be part of the plan.
In 2020, Health Canada and Innovation, Science, and Economic Development Canada launched an advisory council to provide oversight and share guidance on decisions made about the app. From the terms of reference of that advisory group, as well as their first interim report in Feb 2021, part of their job was: “Providing advice for eventual wind-down of the app, including recommendations for the timely destruction of data”. So let’s say this part clearly: as a condition for launch, the government signaled its intent to wind-down the app.
This is important to hold onto because the need to sunset the app is clear — it’s broken — and the government said it would do so. Now, whether that was real or theatrical can be debated. The rest of that interim report from February 2021 suggests a set of plans that were/are in direct conflict with this idea of putting the app away. Here’s an excerpt from near the end of the first report, setting up future reports on these topics that have so far not materialized. Please read the report in full if you’re able:
“The Government of Canada has begun to broadly consider how the COVID Alert app could potentially extend beyond a government service to Canadians and the public health system towards a tool that will also support Canadians and businesses in our economic, social and mental health recovery and restoration. To this end, it will be critical for individuals and businesses in Canada to have trust in the app’s ability to support their safe return to worksites and universities, their reopening of businesses, and their use of modes of transportation including public transit (air, marine, and rail services) until the pandemic is declared over. The advice of the Council will help to inform the Government’s next steps in all of these regards.”
This council has gone quiet since May 2021, so to be sure and fair, these plans may have already been stopped. I am currently making my way through all of the summary reports available on the website and trying to figure out what happened with that now silent council. But the broad intent for expansion of the app was there, and made “public” in this report. And at the moment, the last statement I’ve seen from Health Canada is both vague, technically incorrect, and does not explicitly address any of these plans from 2021. More on that to close.
But before I do, an important point to make note of as a government tactic here for future: part of how the government dealt with critique of this app when it was launched was to commit to shutting it down when it was no longer in use. As such, they have to be held to that idea. Again, this is much more about the future of technosolutionist state interventions and our larger digital public infrastructure than this one particular app.
What It Means in Practice to Support an Application
In this CBC post-mortem piece from February 2022: “Health Canada said that while it continues to support the app, no new features or major updates are planned. It hasn’t received an update since Aug. 9, according to Google’s Play Store. It received its most recent maintenance update on iOS five months ago.”
So here we have another problem. This statement is in that world of true/untrue. Health Canada saying it continues to support the app, on one hand (which in this case is defined as leaving it up) but at the same time, the core function of the app is broken in multiple provinces (because of limited PCR testing capacity). If you download the app, like I did this weekend, you wouldn’t know that there was any such problem. Zero communication of the issue in the app.
This is not what supporting an app looks like in practice. This is evading maintenance and ignoring responsibilities. The post-mortem style of the CBC piece focuses on costs and efficacy, but it let this part of the equation of what it means to “support an app” go unchallenged. If you read the statement it may seem fine. It’s not fine when you bring in other information about how the app functions, information the government absolutely has intimate details of.
There are other issues that tie into this PCR test result requirement that demand further exploration: jurisdiction, the Google/Apple requirement of one-instance per country, and how that looks in practice in Canada given federal/provincial jurisdiction on health, etc. being a big one. But for today let’s focus on what we know from what the government published. Health Canada is the business owner of the app, Health Canada has the following responsibilities as per the terms of reference of the advisory committee:
“Health Canada is the business owner of the exposure notification service with responsibility for engagement with the provinces and territories, privacy and legal, and the communications strategy. The Secretariat support for the Advisory Council will be co-hosted by Health Canada and Innovation, Science and Economic Development Canada.”
Assuming business owner means the one with the responsibility to wind this down and communicate that process, Health Canada has a significant amount of additional communicating to do. There is a story they can tell, some myth if need be, that they tried this, they learned a lot, and now they are doing the responsible thing they said they would always do, which is shut it down.
There is more than sufficient knowledge within the public service on how to write this message, how to work through the process with the public on how to uninstall, and to show that this can all be done. This is part of supporting digital literacy, keeping to open government commitments, and demonstrating public accountability. And it’s a good long-term muscle for all of us to develop together in preparation for next time.
PS: Thanks always to my colleague Sean McDonald for sharing all his lessons learned in this realm. I didn’t get into it here, but when I mention the policy work that needs to get done at the front of this kind of a launch to avoid being where we are today, this is all part of that longer conversation with Sean and many others. Here are a few of the last pieces we wrote together earlier on — First Policy Response, a recent blog post — that inform that thinking and future work.