On April 7, 2026, Canada's Competition Tribunal issued a pair of unglamorous but telling procedural rulings in the Competition Bureau's abuse-of-dominance case against Google. It ordered Google to confirm the accuracy of foreign witness testimony drawn from the U.S. Department of Justice's ad-tech case so the transcripts can be used in the Canadian proceeding, and it rejected Google's bid for what the Tribunal described as a "boatload" of internal Bureau communications. Neither ruling decides the merits. Together they signal something more consequential: the threshold fights are over, and the case is now about substance — and, eventually, remedy.
That shift became possible on March 4, 2026, when the Tribunal dismissed Google's constitutional challenge and reaffirmed that the Bureau's case could proceed. Acting Commissioner Jeanne Pratt said the Bureau continues to "stand by our investigative findings that, through its anti-competitive conduct, Google has been able to entrench its dominance, prevent rivals from competing, inhibit innovation, inflate advertising costs and reduce publishers' revenues."
How the case got here
The Bureau filed its application on November 28, 2024, asking the Tribunal to order Google to sell two pillars of its advertising stack — the DoubleClick for Publishers (DFP) ad server and the AdX exchange — and to pay an administrative monetary penalty. The theory is tying and self-preferencing: that Google links the tool publishers use to sell inventory, the tool advertisers use to buy it, and the exchange that matches them, in order to entrench dominance across the chain. The Bureau pegs Google's Canadian shares at roughly 90% of publisher ad servers, 70% of advertiser networks, 60% of demand-side platforms, and 50% of ad exchanges. Google's vice-president of global ads, Dan Taylor, has called the sector "highly competitive" and said the complaint "ignores the intense competition where ad buyers and sellers have plenty of choice."
The strongest version of the Bureau's case
It deserves to be stated plainly before it is contested. A single firm that operates the buyer's tool, the seller's tool, and the marketplace between them has both the means and the incentive to steer auctions toward its own products and to set the rules in its own favour. The DOJ won a U.S. liability finding on essentially this conduct, and the Tribunal's April decision to import that testimony reflects how much factual overlap exists. When one company controls the ad server roughly 90% of publishers must use, "choice" elsewhere in the stack is a weak constraint. The resulting harms — higher ad costs, thinner publisher margins, less incentive to innovate — fall on Canadian news outlets already under financial strain. A regulator that ignored that concentration would be failing its mandate.
Where proportionality comes in
Granting all of that, the remedy the Bureau seeks is the most drastic instrument in the competition toolkit. Forced divestiture of DFP and AdX would not restore a lost market; it would attempt to re-engineer the plumbing that carries most digital ad dollars in Canada, with no guarantee the spun-off pieces would be more competitive or more useful to publishers. Integration also produces genuine efficiencies — lower latency, fraud detection, unified billing — that a clumsy break-up can destroy along with the abuse.
The better path is conduct-based: enforceable interoperability, auction transparency, a ban on self-preferencing, and data portability so rival exchanges and servers can compete on the merits. These remedies target the mechanism of harm — the steering and tying — rather than the corporate structure, and they can be calibrated and tightened over time. The U.S. case is itself a live experiment in whether structural separation is even workable; Canada would be unwise to commit to the heaviest remedy before that question is answered.
Borrowed evidence, Canadian record
The April 7 evidence ruling is efficient, but it carries a risk worth naming. Canada should not outsource its factual record to a foreign court applying different statutes and market definitions. U.S. ad-tech findings are persuasive context, not a substitute for proof that Canadian publishers and advertisers were harmed in the Canadian market. The Tribunal was right to deny Google's sweeping demand for internal Bureau communications — discovery is not a fishing licence — but the flip side of that discipline is that a Canadian remedy must rest on a Canadian evidentiary record, not a transplanted one.
The C$91 billion question
Google's constitutional challenge was always a long shot, and the Tribunal rightly rejected it. Section 79 of the Competition Act caps an administrative monetary penalty at the greater of fixed amounts or 3% of a firm's annual worldwide gross revenue. Applying that formula to Google's 2008–2024 revenues, Google estimated its own worst-case exposure at roughly C$91 billion. Applying the Supreme Court's Guindon test, the Tribunal (in Commissioner of Competition v. Google, 2026 Comp Trib 10) found the AMP regime regulatory rather than penal — a cap meant to deter, not a criminal punishment triggering Charter section 11 protections.
That is sound law, but "constitutional" is not the same as "sensible." A penalty scaled to global revenue for conduct in a single mid-sized market invites headline numbers that bear little relation to actual Canadian harm. If the Tribunal eventually finds liability, the discipline it showed on discovery should extend to the penalty: an amount proportionate to the harm proven here, paired with forward-looking conduct remedies, would do more for Canadian publishers and the open ad-tech market than a record fine or a speculative break-up. Proportionate enforcement is not soft enforcement — it is the kind that survives appeal and actually changes behaviour.