Will Canada and the European Union sign a Digital Trade Agreement by December 31, 2027?

TL;DR

I assign a 77% probability of YES. Four fast negotiating rounds, a reported Commission target to conclude at the October summit, and the EU trade commissioner’s public reference to possible 2026 signature show a deliberate push to beat the normal timetable (MLex; Euronews). The main failure mode is delay rather than collapse: if substantive negotiations extend past the first quarter of 2027, the EU’s legal and Council process can push signature into 2028.

Context

Canada and the EU formally launched negotiations on March 5, 2026. By the September 22 cutoff, the Commission had published reports for four rounds—April, June, July, and August—plus its textual proposal (Commission document index). The question’s primary resolution page still described the Canada talks only as negotiations that began in March, not as concluded or signed (European Commission).

This is a timing forecast for formal signature. Ratification and entry into force are not required. The Commission’s negotiating recommendation places the intended agreement within the EU’s exclusive common-commercial-policy competence, making an EU-only route the base case (Commission recommendation). Canada likewise handles signature through Cabinet approval and an Order in Council; parliamentary tabling and implementation normally follow signature (Canadian treaty process).

Evidence

The closest reference class contains only two completed standalone bilateral EU Digital Trade Agreements, observed from their launches through June 10, 2026. Singapore ran from July 20, 2023 to July 25, 2024 for negotiation, then to May 7, 2025 for signature: 371 calendar days to conclusion, 286 more to signature, and 657 in total (Commission chronology). South Korea ran from October 31, 2023 to March 10, 2025 for negotiation, then to June 10, 2026 for signature: 496 days to conclusion, 457 more to signature, and 953 in total (Commission chronology). Canada has 666 days from launch to deadline. A mechanical base rate is therefore 50%: Singapore fits by nine days; Korea misses by 287.

The post-negotiation phase is less random than the final signature dates suggest. Singapore took 263 days from conclusion to Council authorisation, while Korea took 259. In both cases, roughly six months passed before the Commission submitted its signature proposal, followed by about two and a half months of Council processing (Singapore procedure; Korea proposal; Korea Council decision). The variable part was ceremony timing: Singapore signed 23 days after authorisation, while Korea waited 198 days for a summit (Singapore Council release; Korea summit).

Canada is ahead of that raw base rate. Round 1, held April 14–16, identified convergence and gaps (round-one report). Round 2, held June 16–18, reported progress on several provisions (round-two report). Round 3, held July 21–24, reported substantial progress on data flows, privacy, electronic contracts and other disciplines (round-three report). Round 4, held August 24–26, was the first to say that negotiators had agreed in principle on mutually acceptable textual provisions where possible, while reporting good progress on data flows, source code, electronic invoicing and duties on electronic transmissions (round-four report). A fifth round was being arranged in Brussels.

The talks are not finished. Round 4 still listed scope, personal-data protection, cooperation, online consumer protection, open internet access and open government data as under discussion (round-four report). Privacy is a real last-mile risk: European civil-society groups asked EU institutions on September 14 to raise Canada’s proposed lawful-access legislation during the DTA negotiations, though no official EU source had made that legislation a veto (Access Now letter).

The strongest live signal is an August 19 report that a Commission document set a goal of concluding negotiations by the October 29–30 Montreal summit, with financial-data rules, scope and source code named as outstanding issues (MLex; official summit page). On September 2, Trade Commissioner Maroš Šefčovič went further and said a digital agreement might be signed before the end of 2026 (Euronews). The wider relationship was then elevated on September 16, when Commission President Ursula von der Leyen proposed moving from CETA to an Alliance for the Future that includes a technology alliance and common prosperity space (State of the Union address). This creates a strong incentive to make the DTA an early, concrete deliverable.

My timing model is:

The weighted result is 0.58×0.89 + 0.22×0.83 + 0.12×0.58 + 0.05×0.10 + 0.03×0.01 = 0.7737. The first column reflects the fast rounds and summit target, discounted for the issues still open. The conditional probabilities reflect the roughly 260-day conclusion-to-Council-authorisation period in both completed EU DTAs, plus uncertainty over how quickly the parties arrange signature afterward.

What's non-obvious

The obvious 50% base rate is too low because it treats Canada as if it were following the same negotiation clock as Singapore and Korea. Canada completed four rounds in less than six months, versus five Singapore rounds over October 2023–May 2024 and six Korean rounds over December 2023–October 2024 (Singapore round history; Korea round history). An October or December conclusion would buy several extra months for the otherwise slow EU legal process.

Šefčovič’s suggestion of a 2026 signature is too aggressive to take literally as the central case. It would require a large break from both prior post-conclusion timelines. I read it instead as evidence that political principals expect rapid substantive closure and may already be pushing legal review, translation and approval work in parallel. Signature also avoids the national-ratification trap associated with CETA: neither all 27 national parliaments nor European Parliament consent should be required before the signing act itself (EU mandate; Singapore Council precedent).

Uncertainties

The October target rests on a non-public Commission document reported by MLex, not on a public commitment. None of the four round reports gives a conclusion or signature date. The commissioner may also have used “signed” loosely when he meant politically concluded.

The reference class has only two standalone agreements, so ceremony timing and legal-processing variance are poorly estimated. The public material does not show how much legal scrubbing or translation has already occurred in parallel, nor whether Canada’s lawful-access legislation will become an official EU negotiating issue. A formal conclusion at the October summit would move the forecast close to 90%; a summit that produces only a general progress statement would reduce it to roughly 60–65%.