How much Canadian research is free to read, how it got there, and how ready tri-agency-funded papers are for the 2026 Tri-Agency Open Access Policy, which applies to grants awarded from 1 January 2027. Built on open data from OpenAlex. Think of it as the starting line, not a compliance report: papers published so far fall under the 2015 rules.
The agencies published two KPIs on 5 October 2026 and will track them "using open data sources"; I couldn't find their method. Ours, from OpenAlex, run higher: about 6 points for Canada and 5 to 10 for funded papers. For funded papers the likely reason is how funding is detected: for closed papers it comes mostly from what publishers deposit with Crossref, while OpenAlex can also find funders in full text, which it mostly has for open papers. Both count papers that are free to read anywhere, which is not the same as meeting the new policy.
The Canadian repository line falls after 2019. Some of that is lag, since deposits often come after publication. Some is harvesting that is still being fixed (see Repositories below). Some may be authors no longer depositing papers that are already open through read-and-publish deals, which would be a finding in its own right. Which of these dominates is being checked.
Tri-agency papers published in 2025, set against what the policy will ask of papers from grants awarded from 2027.
Share of each agency's papers whose metadata carries that agency's grant number. The 2026 policy makes the grant number a requirement, so this should climb toward 100%.
Tri-agency papers from 2025 with no free copy, by publisher, coloured by whether the publisher lets authors share the accepted manuscript at once, free, under an open licence. Publishers' public policies as of 18 Nov 2026; the agencies' letter to publishers is still open.
Share of tri-agency papers from 2025 that are free to read, by the paper's main field (fields with 200 or more papers).
Share of each funder's 2025 papers that are free to read. Funders' papers are found in different ways (NIH's largely through PubMed grant lists, most others through publisher metadata), so read small gaps with care.
The repository numbers are only as good as the harvesting behind them, and Canada has made that harder than it needs to be. France runs one national repository, HAL, that every institution deposits into and every librarian can help curate. Canada took a hybrid route: many universities run the same software through Scholaris, the shared service from Scholars Portal, but each has its own instance and its own harvest address, which changes when a repository moves.
In September, OpenAlex audited the 26 live Scholaris repositories. Only 2 were being harvested, because most had migrated and we were still reading their old addresses. We repointed and registered all of them, and today all 26 are harvested, about 595,000 records. In a test set of 392 known deposits, the number we could find went from 195 to 242. A few repositories outside Scholaris still show sudden drops in recent years, and those are being chased now.
We're committed to keep working with the Scholaris community so OpenAlex indexes as much of Canada's repository content as it can. Refinements, like a view by university, will follow as the harvesting does.