It started with an unexpected early-morning announcement on June 3, 2026, from Marc Miller, the Minister of Identity and Culture. Mr. Miller said that the government planned to direct the Canadian Radio-television and Telecommunications Commission (CRTC), Canada’s broadcast regulator, to review its two-week-old decision that imposed hundreds of millions in new investment requirements on internet streaming services. My Globe and Mail essay that appeared over the weekend notes that the move came as a surprise, not only because he had chastised the commission a month earlier for moving too slowly, but also because it marked a major reversal of a core Canadian digital policy that had been years in the making. The decision sent shock waves through the cultural sector, but it was only the start.

Over the two weeks that followed, the government unveiled its artificial intelligence strategy, introduced a social media ban for anyone under the age of 16 and a new superregulator for social media and AI chatbots, brought forward a privacy reform bill that removes the Privacy Commissioner of Canada from regulating private-sector privacy, and rushed through lawful access legislation that left leading tech companies contemplating leaving Canada altogether. Those focused on digital policy are left struggling to make sense of the dizzying pace of change and a digital strategy seemingly in disarray.

Making sense of it means stepping back, because those two weeks are best understood as digital strategy 3.0 for the Liberal government. Version 1.0 was a warm embrace of technology, viewed as central to a new government’s vision of a progressive, innovative country. The policy was largely hands-off, without significant new regulation. Indeed, the government was more anxious to demonstrate its openness to tech: Google’s smart-city plan for Toronto’s waterfront drew then-Prime Minister Justin Trudeau to the launch announcement, and then-heritage minister Mélanie Joly travelled to Silicon Valley, returning triumphantly with a five-year deal committing Netflix to $500-million in investment in Canada.

Version 2.0 brought an end to the cheerleading. As the downsides of tech became more apparent, the government reversed course, pursuing a trio of laws to lay the foundation for a more aggressive regulatory approach. The bills that became the Online Streaming Act, the Online News Act, and the failed Online Harms Act were sold as ushering in new regulation but were better captured by the slogan “make web giants pay.” The government seemed less interested in constraining the behaviour of tech giants than in benefiting from it by requiring these companies to help fund television and film production, news, and the regulatory model itself. To those payments, it added an expanded taxation approach that required digital services to collect and remit sales tax and to pay a new levy on revenues in the form of a digital-services tax.

By the time Mark Carney became Prime Minister, the failures of the “make web giants pay” approach were obvious. News links were blocked for years on Facebook and Instagram, leading to fewer news choices online and only limited additional revenue for news organizations. The streaming act was stuck in the courts and in lengthy regulatory hearings, owing to the lack of detail the government had placed in the law. Other initiatives failed altogether: The Online Harms Act died, as did a poorly constructed privacy bill that included a bare-bones plan to regulate AI with virtually no specifics.

All of this brings us to Version 3.0. It really started a year ago, with the decision to drop the digital services tax under trade pressure from the United States, though it had become apparent that Mr. Carney needed little encouragement to move on from his predecessor’s approach. Even so, the two weeks that just passed are hard to fully process.

The streaming reversal is the real end of Version 2.0, with the government deciding to direct the CRTC for a second time on how it wants its own law interpreted. The AI strategy arrived months later than promised, with the government focused on adoption through multiple new funded programs and a commitment to developing new rules to foster trust. Those rules came quickly but seem unlikely to achieve the trust objective soon. Bill C-34, the Safe Social Media Act, creates a duty to act responsibly for social media and AI companies, but that sensible foundation is largely lost amid the risks of a social-media ban that is dominating the discussion, the uncertainty of at least 50 issues left to Cabinet or a new Digital Safety Commission to decide, and the years it will take to get the entire regulatory edifice off the ground. The government doubled down days later with Bill C-36, a privacy reform bill that contains many modernized provisions but relies on a regulatory framework that sidelines the Privacy Commissioner and will take even longer to implement. Finally, there is lawful access, which the government pushed through by cutting off committee debate despite privacy and security risks that have sparked widespread alarm.

It is tempting to say there isn’t an underlying policy with siloed ministers working at cross-purposes. How else to explain a government that says it wants legislation declaring privacy as a fundamental right, but passes rules such as mandated metadata retention that represent some of the biggest privacy threats in years? Or a government that insists on a privacy review of age verification technologies to be used in a social-media ban in Bill C-34, only to reverse course five days later by repealing the privacy safeguard in Bill C-36? These moves, alongside efforts to woo AI companies with one hand through the AI strategy while driving them out with the other through lawful access, do not inspire confidence.

But look carefully, and there are two themes — one that seems to underlie the policy, and another designed to sell it. The theme underlying the policy is that Canada’s struggles with the digital world are fundamentally a governance problem. There are odd governance choices, though they tend to sit at the ministerial level, where the privacy rules that have long governed banks, retailers and other businesses are now the domain of Evan Solomon, the AI minister, and where public safety rules for social media are Mr. Miller’s responsibility.

Yet the government is not fixing its own leadership structure. Instead, outdated privacy regulations and the absence of strong penalties are apparently the Privacy Commissioner’s fault, and delays in streaming rules are blamed not on poorly constructed legislation but on a slow-moving regulator. The governance solutions may well make things worse, however. The government is betting big on more centralized power over policy decisions and on a new superregulator responsible for all things digital, which will take years to become fully operational.

Governance is a hard sell, however, so the government is pitching this as a way to protect children. That has led to a consistent framing of digital policy in which breaking encryption may be needed to enable law enforcement to save kids from imminent harm, social-media access should be banned for those under 16, and unique privacy rules are needed to protect sensitive information involving children.

Kids may be a reliable way to frame policies in need of support, but there is ample reason for skepticism. Bans and backdoor access to security systems have not worked elsewhere and are likely to cause their own harms, including weakened security, privacy risks that accompany requiring tens of millions of Canadians to verify their age to use social media, the reputational damage to Canadian privacy from the reduced role of the Privacy Commissioner, and the potential exodus of leading tech companies that provide essential security services. If these come to pass, it will be Canadian kids who ultimately pay the price for the digital policy missteps of two weeks in June.