We surveyed every AI governance vendor selling into Canada. Nine of twelve will not show a price before a sales call, and none of them covers Law 25, OSFI E-23, and Bill 194 natively. Here is what we charge, what they charge, and what doing nothing costs.
Start with a paid 90-day Inventory Sprint that maps every AI system you run. Then the platform is priced by institution size, with unlimited AI systems monitored: your price never moves as you add models, unlike the per-model or per-seat pricing of the horizontal incumbents. Entry starts below IBM's published price, in Canadian dollars. Founding design partners are the one exception to size-based pricing: a flat $9.6K a year for two years, whatever their assets.
A fixed 90-day diagnostic, priced by your institution size and credited in full to your first platform year.
The continuous platform, priced on your total assets, not your AI count, so your price holds as you grow.
For the largest institutions, and any firm that needs private deployment, deeper controls, and dedicated support.
Everyone else pays by size. The first twelve institutions to build Veille with us don't: a flat $9.6K a year for two years, whatever they're worth. You help shape what we build, with a direct line to the founders. After that, you move to your standard rate.
Compiled and re-verified in July 2026 from public marketplace listings, procurement data, and vendor pages. Where a vendor hides its price, we say so, because that is data too. A green check means the framework is covered natively.
| Solution | Annual cost | Pricing model | Law 25 | AMF Québec | OSFI E-23 | Notes |
|---|---|---|---|---|---|---|
| Veille | From $12K CAD | By institution size, unlimited AI systems, published | Built in Montréal, bilingual, data in Canada. The only platform that encodes the AMF guideline and Law 25 as tracked obligations, not only OSFI E-23. | |||
| ValidMind | Quote only | Per deployment, on request | Model risk | Purpose-built model-risk platform with a named Canadian bank client. Strong on OSFI E-23, nothing on Québec privacy or the AMF. US-based. | ||
| Collibra | ~US$170K+ (est.) | AI module plus platform base, on request | Privacy config | Blog only | Horizontal data-governance suite. Maps to E-23 in a blog, not a maintained control set. No Law 25 or AMF product. | |
| IBM watsonx.governance | US$38K (~$52K CAD) | Per instance, covers 5 AI use cases, then US$15,960 per added case | Generic MRM | The only public list price among the global vendors. Fifteen use cases approach US$200K per year before integration. | ||
| OneTrust | US$50K to $300K+ | Per module plus admins, on request | Privacy side | The strongest foreign vendor on Law 25, but on the privacy suite, not AI governance. AMF and E-23 absent. | ||
| RegCore.AI | Not published | Advisory plus playbooks, no self-serve product | Named | Named | Toronto firm bundling Law 25 and E-23 for financial institutions. Advisory-led, not a self-serve platform, and missing the AMF. | |
| VerifyWise | Free, open-source | Self-hosted, or managed on request | Listed | Blog only | Open-source. Lists Law 25 among 24-plus global frameworks; E-23 is written content, not encoded controls. No AMF. | |
| Big 4 advisory | US$300K to $500K | Hourly, with initial builds from US$500K over 18 to 24 months | Project | Project | Project | Point-in-time deliverables. The report is out of date the day the regulation changes. |
At CAD $12K, Veille enters below IBM's published list price (US$38K, about $52K CAD, the only public price in the market) and below the reported ranges of Collibra, OneTrust, and Big 4 advisory. It also sits under the category's own floor: independent 2026 analyst benchmarks put enterprise AI-governance platforms at US$50K to $200K per year. The line that matters is coverage, not price: several vendors now map OSFI E-23, none map the AMF Québec AI guideline, and none encode Law 25 and the AMF together as tracked obligations. Sources: AWS Marketplace listing (IBM watsonx.governance, re-verified July 2026), vendor pricing and documentation pages, third-party procurement benchmarks (Vendr and Spendflo, 2025 and 2026), and analyst estimates where marked as such. ValidMind, OneTrust, and RegCore.AI do not publish list prices; the Collibra figure is a third-party benchmark estimate, not a vendor-published price. VerifyWise is open-source and free to self-host. USD converts at roughly 1.37 to CAD, a conservative rate: the live July 2026 rate is closer to 1.42, which only widens the gap. Framework coverage reflects each vendor's published materials as of July 2026.
Most organizations do not buy a competing product. They buy hours. Here is what those hours cost in Canada, sourced in June 2026.
A typical "do nothing new" basket, fractional counsel plus external assessments plus one compliance hire, runs past $150K to $250K CAD per year, refreshed annually, with no continuous record. The Veille platform starts at $12K CAD, with unlimited AI systems monitored, below IBM's published entry price.
These are the statutory maximums in force today, not projections. Enforcement capacity at the CAI has been expanding since 2023.
Founding design partners take a bet on an early-stage product and shape it with us. In exchange, the first twelve institutions pay a flat $9.6K a year for two years, whatever their size, even a large institution that would otherwise sit in a higher band. It is the one exception to size-based pricing, and a fixed two-year term, not a permanent discount: when it ends, you move to your standard size-based rate.
Nothing beyond the sprint. In 90 days we find every AI system you run, shadow systems included, map the gaps against the AMF guideline, Law 25, and OSFI E-23 if federally regulated, and hand you a posture report. If it does not prove its value, you keep the report and walk away. If you continue, the sprint fee is credited in full to your first platform year.
Because most organizations do not know how many AI systems they run. Shadow AI is, by definition, the AI you have not found. Charging per declared system would reward under-declaring the very shadow AI we exist to surface. So we price on institution size, total assets, which is public and ungameable, and we monitor unlimited systems. The more we find, the more complete your coverage, and your price stays the same.
Yes. Every vendor in the comparison table bills in USD. At current rates that adds roughly 37 percent to their sticker price. We bill Canadian organizations in Canadian dollars, hosted on Canadian infrastructure.
No. The rule library is the product. When a federal AI bill is enacted, or OSFI updates E-23, the obligations are encoded into your existing subscription. Unlike per-framework pricing or per-use-case pricing, your cost does not grow with your exposure.
No, and we put that in writing. Veille is compliance tooling, not legal advice. It makes your counsel far more efficient: instead of paying $900 per hour for document drafting and obligation tracking, you pay for judgment on the questions that genuinely need it. See our Terms.
Start with a 90-day Inventory Sprint against your real AI systems, shadow AI included. Your E-23 and Law 25 posture is live from the report.
Book an Inventory Sprint