The Future Tense

Read the same announcement twice — once for what it funds, once for what it promises — and a grammar appears. On 4 June 2026, launching “AI for All,” the government described how it would protect Canadians from the risks of artificial intelligence: it will modernize privacy law, it will introduce an online-safety regime, it will guard against deepfakes and surveillance pricing. Every protection is a verb in the future tense — legislation not yet written, “vague on a timeline,” with no funding attached. Meanwhile the things that are real today — the adoption fund, the compute money, the AI “missions,” the data-integration contracts the same government is already signing — are in the present tense, costed and moving. This case is about that asymmetry: protection deferred to a bill that may come; capacity delivered in a budget that already has.

§01 — The move named

The temptation is to read a national AI strategy as either a gift or a betrayal and to argue about which. That argument is real, and it is not the one this series runs. Strip the verdict away and a familiar grammar remains: the protection is announced as a plan — the device named in Case 25 · The Same Facts, where a council’s four directions were all process and none a destination — and the contested instruments are deferred to where they are not yet binding, the move named in Case 26. Put the two together and you get a laundering that runs on verb tense: the thing that reassures you is in the future; the thing that should worry you is in the present; and the announcement reads them in the same breath, as if they were the same kind of thing.

Name the structure before the evidence. A strategy can move money this fiscal year. It cannot pass a law this fiscal year — laws are drafted, tabled, studied, amended, voted. So when a plan funds adoption now and promises protection later, the asymmetry is not an accident of messaging; it is the shape of how the two halves actually arrive. This is the direct sequel to Case 31: that case lowered the legal threshold to reach the data; this one funds the capacity to use it, while promising to protect you from it.

• P1 AI has real upside and real risk; a national strategy, public investment, even a health “mission,” can be legitimate and genuinely intended. Grant it in full (§05).
• P2 The strategy funds adoption, compute, and capacity now (e.g. a $700M Compute Access Fund); the protective instruments — privacy modernization, an online-safety regime, chatbot rules — are forthcoming legislation, “vague on a timeline,” with no funding attached.
• P3 Announcing a protection does not deliver it. A line in a budget and a clause in a bill-not-yet-written are not symmetric: one is a fact, the other is a plan.
• P4 The strategy presents “protecting Canadians from AI risk” as the headline while the protection sits in the future tense and the capacity sits in the present tense — and offers the protective vocabulary as if the two arrive together. The asymmetry of tense is the laundering.

Counter: when a plan protects you in the future and builds in the present, read the verbs. Ask what is funded this year, and what is merely promised for a bill that has not been written.

§02 — The present tense

Start with what is real today, because money is the least ambiguous part of any announcement. “AI for All,” launched at Toronto’s University Health Network, is a multi-billion-dollar strategy: the government projects on the order of $200 billion of economic growth and 250,000 new AI-related jobs over five years, and the package includes concrete, costed instruments — among them a $700 million Compute Access Fund to put compute in the hands of businesses adopting AI, work-placement programs for up to 90,000 people, and a flagship health “mission” to accelerate AI in diagnostics and care.primary

These are present-tense things. They have dollar figures, delivery vehicles, and a fiscal year. Note, too, the strategy’s own stated purpose, in the government’s framing and the coverage of it: to close an “adoption gap” and build public trust — that is, to get more AI deployed, faster. There is nothing hidden in that; it is the explicit goal. Hold it in mind, because it tells you which half of the announcement is load-bearing. A strategy whose purpose is adoption funds adoption.

Counter: the present tense is where the commitments are real. Follow the money to the verb that has already happened — “allocated,” “contracted,” “launched” — and you have found what the plan is actually for.

§03 — The future tense

Now the other half — the protection. It is not absent; it is announced. The strategy says the government will modernize consumer privacy legislation (to “enshrine a right to privacy” and safeguard children’s information), will introduce an online-safety regime “to better protect social media and chatbot users,” and will guard against “harmful practices such as deepfakes and surveillance pricing.” Every one of these is a verb in the future tense — “we will,” “will introduce new legislation.” And the reporting that read the documents found the protection thin where the funding was thick.primary

While it proposes new consumer privacy legislation to enshrine a right to privacy and safeguard children’s information — along with a modernization of safety laws — the plan is vague on a timeline and what specific measures will be put forward to do so.BNN Bloomberg, 4 Jun 2026 — “AI strategy pledges thousands of jobs, lacks safety details”

The cleanest specimen is the chatbot rule that did not make it in. Proposed restrictions on chatbots for under-sixteens were left out, the strategy explained, because there is “a current review to determine whether they should be integrated into online harms legislation set for later this year.” That is the future tense doing its work in a single clause: a protection for children, present as an intention, deferred to a bill that does not yet exist. And the criticism was on the record and across the aisle — the Conservative deputy leader, reading the same documents, said the “safety and the security that was promised in this is nowhere to be found in the documents. Certainly, no details.”primary

Every protection is a verb that hasn’t happened yet. Every capacity is a verb that already has.

Counter: “we will protect you” is a promise; “we have funded this” is a fact. A plan that mixes the two in one sentence is asking you to bank the promise at the value of the fact.

§04 — The tell: the adoption gap

The tell is not a critic’s; it is the strategy’s own stated purpose. The plan describes its job as closing the “adoption gap” and building public trust. Read those two together and the relationship between the halves is named by the document itself: trust is instrumental — it is the thing you build so that adoption can proceed. Protection, in that frame, is not the destination; it is the on-ramp. The privacy bill and the safety regime are how you make people comfortable enough to accept the capacity that is already funded.

The Prime Minister’s own line carries the conditional that gives the game away: AI, he said, can shorten emergency-room waits and make a small business more competitive — “if it is governed by Canadian values.” The promise rides on a conditional, and the conditional — the governing — is exactly the part deferred to legislation not yet written. The benefit is offered in the present; the “if” is filed for later. This is the integration stamp of Case 25 with a clock attached: the safeguard announced as a plan, then post-dated.

Counter: when a government tells you the benefits are conditional on governance, and then funds the benefits while deferring the governance, the conditional is decorative. Ask for the date the “if” becomes law.

§05 — The strongest version of the other side

Build the government’s case at its strongest, because it is not weak. Laws genuinely cannot be passed in a budget cycle; tabling a serious privacy bill or an online-harms regime takes drafting, consultation, and a parliamentary calendar, and a responsible government may legitimately fund adoption while that work proceeds. The protections named are real proposals, not nothing — a right to privacy, protections for children, rules against deepfakes and surveillance pricing are worth wanting. Investment in compute and skills can serve the public. And “we will legislate” is the honest grammar of a measure that has not yet been written; it is not, by itself, a deception.analysis

Grant all of it. The narrow claim survives, because it was never that the protections are insincere or that investment is wrong. It is that a funded capacity and a promised safeguard are not the same size, and that announcing them together invites you to treat them as if they were. The honest version of this strategy would say: here is what we are building now, and here is the protection that does not exist yet, and here is the date by which it will. The laundering is in the missing date — in letting the future-tense protection borrow the solidity of the present-tense spend. Sequence matters: when capacity ships first and governance ships “later this year,” the gap between them is a period during which the capacity operates ungoverned.

Counter: the strongest case for the plan is also the test of it. If the protections are real, they have a tabling date. Ask for it; a sincere safeguard can name its own deadline.

§06 — The specimen: procurement doesn’t wait

Here is where the present tense is most concrete, and where the case is most careful. The same government that promises a future “data sovereignty” framework is, now, holding live data-integration procurement. Reporting by the Investigative Journalism Foundation and The Logic has documented a Canadian Defence contract with Palantir — first signed in March 2020 at about $14.4 million and amended upward to roughly $44 million by October 2025, with Defence spending on it reported around $46.8 million — held with the special-forces command, plus a separate 2026 “data integration and analytics platform” call-up. The minister responsible, David McGuinty, called the latest deal “a legitimate procurement” that is “moving forward,” and said the government “will look at this question of data sovereignty” — a sovereignty framework that, as of the launch, is pledged but not yet enacted.primary

Note carefully what this is and is not. It is not a claim that the $2.3-billion strategy is Palantir money, or that the contracts are a secret surveillance program; the procurement is a separate, disclosed line, and merging the dollar figures would be the very overreach this series refuses. It is the asymmetry in its sharpest form: the capacity to integrate data is being bought in the present tense, while the instrument that would govern it — sovereignty, the “if it is governed by Canadian values” — is promised in the future tense and not yet enacted. That is the spine, in procurement.

The thread that pointed here read the same facts and reached louder conclusions: that this is a “permanent surveillance state,” that the strategy is an industry “bailout,” that dissent will be recast as a security threat. Those are real positions, argued by real people, and they are carried here as named lenses, attributed and not adopted — alongside the government’s opposite lens, that this is responsible, sovereignty-minded, privacy-first modernization. The series adjudicates neither. It only fixes the tense: capacity now, governance later, and a gap between.analysis

Counter: the test of a sovereignty promise is whether it constrains a contract you can already sign. A safeguard that does not yet exist cannot govern a capacity that already does.

§07 — The future tense, named

Strip it to the structure, and no villain is required. A government has a real opportunity and a real risk, and it announces a plan for both. The opportunity it funds: compute, adoption, jobs, a health mission, procurement — present tense, costed, moving. The risk it addresses by promising to legislate: privacy, online safety, deepfakes, the rules for chatbots and children — future tense, undated, unfunded. Each half is announced in the same speech, in the same protective vocabulary, as if both were already true. The integration is a citizen invited to bank a promise at the value of a fact, and to accept the capacity today on the security of a safeguard scheduled for later.

That asymmetry is the laundering, and naming it requires adopting none of the louder claims around it. The protections may well arrive; if they do, on a real timeline, the case for this plan strengthens and the gap closes. The series’ insistence is narrow: that the announcement be read in the tense it was actually written in — that “we will protect you” not be allowed to settle the bill that “we have funded this” opens. It sits beside Case 31’s lowered threshold, Case 25’s plan-as-stamp, Case 26’s deferral, and Case 27’s use of the calendar as a method.

When the next plan protects you in the future and builds in the present, do the one thing the grammar is built to discourage. Separate the verbs. Put what is funded this year in one column and what is promised for a bill not yet written in the other, and ask why the protection is always the column without a date. If the safeguard ships with the capacity, the plan is honest. If the capacity ships now and the safeguard ships “later this year,” you are not reading a protection. You are reading a postponement.

Capacity in the present tense. Protection in the future tense. The gap between the verbs is the case.