The rumour has been circulating in Gastown coffee shops and Mount Pleasant co-working spaces for months: a fund called Cascadia VC is writing $1M+ cheques to pre-revenue AI startups in Vancouver right now. It is a good story. It is also, based on every primary source available, unverified.
That gap between the rumour and the registry is the actual story.
The Fund That Isn't — And the Registry That Would Know
Start with what can be confirmed. The closest real entity to "Cascadia VC" is the Cascadia Venture Forum, a Vancouver-based firm established in 2016 with a portfolio of roughly 45 companies. According to Tracxn data, its average seed round sits at approximately $825K. Its primary focus is Life Sciences and HealthTech. It is not an AI-focused fund, and it is not the fund founders are whispering about.
The BC Securities Commission maintains a public registry of every fund registered or operating under an exempt-market dealer exemption in the province. Any entity writing $1M+ cheques in BC must appear there or be in violation of securities law. A search of that registry does not surface a fund operating under the Cascadia VC name with an active AI mandate. That is not a minor footnote. It is the story.
Vancouver has a long history of funds and forums sharing overlapping regional branding — Cascadia, Pacific, Northwest. First-time founders, in particular, regularly conflate networking organizations with active fund managers. The confusion is not innocent: it shapes where founders spend their pitch energy, and pitch energy in a pre-revenue company is a non-renewable resource.
Canada's AI Seed Market Is Smaller Than One SF Series B
Here is the number to hold in your head for the rest of this article: $24M CAD.
According to the CVCA's September 2025 report, "The Current State of Seed Investing in Canada," AI was the top pre-seed and seed vertical nationally in H1-2025, securing $3.9M at pre-seed and $20.1M at seed — $24M total across 12 deals. That is the entire country. That is also a number a single mid-sized Series B in San Francisco would absorb without noticing.
BC attracted $28.5M CAD in total seed-stage investment in H1-2025 across all sectors, per the same CVCA report, ranking fourth provincially. The AI slice of that BC figure is not broken out separately, but the national context makes clear that anyone describing Vancouver as a serious AI funding hub needs to hold that $24M figure in view at all times.
The broader market: according to CVCA's Q3 2025 VC and PE Canadian Market Overview, published November 2025, Canada saw $4.9B invested across 386 VC deals year-to-date through Q3, with an average deal size of $14.7M. That average is distorted heavily by late-stage rounds. At the seed and pre-seed layer, the numbers are structurally thin — and in BC specifically, the pre-revenue AI cheque-writing market is thinner than the national conversation suggests.
Federal Money Is Not a 2026 Funding Solution
Federal Budget 2025 committed two large figures that founders keep misreading as near-term liquidity. Finance Canada allocated $925.6M over five years for sovereign AI compute infrastructure and $1B over three years via BDC's Venture and Growth Capital Catalyst Initiative. Neither of these is a direct-to-startup program.
The $925.6M is infrastructure spend — compute clusters, national research capacity. It does not write cheques to a pre-revenue team in Gastown. The $1B Catalyst Initiative flows through fund managers as LP capital. That means the deployment timeline to an actual founder is 18 to 36 months minimum: fund formation, capital calls, deal sourcing, term sheets. Founders treating Budget 2025 as near-term runway are making a serious planning error.
InBC Investment Corp — BC's government-backed VC, which had committed $140M across 12 direct company investments and 7 venture fund partnerships by end FY2024-25, per its Annual Report — explicitly does not fund pre-revenue companies directly. That is a deliberate policy choice rooted in the BC government's mandate to deploy capital alongside private co-investors, not replace them. The practical effect: the province's most visible VC instrument creates a floor that sits above where most pre-revenue AI startups actually are. Founders who pitch InBC at the idea stage are burning relationship capital they will need at Series A.
The SR&ED program — which provides over $4.2B annually in R&D support to Canadian businesses, with Budget 2025 adding $440M per year in ongoing enhancements per Finance Canada — remains the primary non-dilutive support mechanism for pre-revenue AI startups in BC. Most first-time founders in this cohort are leaving that money on the table. Their accountants are generalists, not SR&ED specialists, and they are misclassifying experimental AI research as product development. That misclassification is expensive.
The Actual Pre-Revenue Capital Stack in BC Right Now
The real funding structure for a pre-revenue AI company in Vancouver in 2026 is a three-layer arrangement that most founders are not mapping correctly.
At the base: SR&ED tax credits, both federal and provincial. BC's Scientific Research and Experimental Development Tax Credit stacks on top of the federal program. The BC Interactive Digital Media Tax Credit — a 17.5% refundable credit on eligible BC labour costs — has been used creatively by AI companies with consumer-facing products. For a pre-revenue AI startup burning $150K per month in engineering salaries in Vancouver, the combined federal and provincial non-dilutive recovery can represent $40K to $60K per month in effective cash return. That is a full quarter of additional runway without touching equity. Most founders in this cohort are not claiming it.
Above that: angel syndicates and family offices writing $250K to $750K cheques. This is the most active layer in the BC pre-revenue AI market right now, and it is almost entirely relationship-driven and invisible to national data aggregators.
At the top: institutional seed rounds from the handful of actual BC-based seed funds. Yaletown Partners and Vanedge Capital — both InBC fund partners — have historically operated above this range at the Series A level. The seed layer in BC is genuinely thinner than the ecosystem narrative suggests, and that $28.5M in total BC seed investment in H1-2025 is spread across all sectors.
One cross-border dynamic deserves serious attention. US investor participation in Canadian pre-seed rounds dropped from 23% to 8.5% in H1-2025, per CVCA data. Three years ago, the standard playbook was: build traction in Vancouver, get a US lead, use that to anchor a Canadian syndicate. That playbook is broken. Whether it is tariff anxiety, a stronger domestic LP base finally emerging, or US funds pulling back from cross-border legal structuring — the effect is the same. BC founders are raising in a more domestically-dependent market than they have seen in a decade, which compresses valuations and extends timelines simultaneously.
The second-order effects of this environment are already visible to anyone watching closely:
- Pre-revenue AI valuations are compressing as founders discover the $1M cheque market is thinner than rumoured.
- SR&ED claim sophistication is accelerating among Vancouver AI teams as non-dilutive capital becomes the primary bridge.
- Cross-border AI startups are dual-incorporating earlier, hedging against domestic capital constraints.
- Angel syndicates are formalizing as micro-funds to access BDC Catalyst Initiative LP allocations.
- Founders with US revenue traction are being rewarded disproportionately, widening the gap between pre-revenue and early-revenue valuations.
Vanhub Intelligence: Local Impact Analysis
According to recent market trends in Metro Vancouver, the tech employment base — particularly in the Mount Pleasant, Yaletown, and Gastown corridors — is structurally sensitive to seed-stage funding velocity in ways that broader labour market data does not capture cleanly. When pre-revenue AI startups cannot close their seed rounds, the first thing that stops is hiring. Not layoffs — hiring freezes. The result is a shadow unemployment effect among mid-career engineers and ML specialists who are between roles and waiting for the next wave of funded companies to open headcount. Vancouver's tech labour market in early 2026 carries more of this latent slack than headline numbers suggest, and a funding gap at the pre-revenue stage is a direct upstream cause.
The real estate dimension is less obvious but worth tracking. The Burrard Corridor and the areas around Great Northern Way — where most of Vancouver's AI and deep-tech startups cluster — have seen commercial lease absorption slow considerably from the 2021-2022 peak. When seed-stage companies cannot raise, they do not sign leases. They work from shared desks and founder apartments. Metro Vancouver operators should note that the office submarket most exposed to this is not downtown Class A towers — those are a separate story driven by law firms and financial services — but rather the 2,000 to 8,000 square foot creative-office and flex-industrial spaces in East Vancouver and Mount Pleasant that were purpose-renovated for tech tenants between 2018 and 2022. Vacancy in that specific submarket is quietly elevated, and landlords are already feeling the shift in tenant negotiating posture.
For Vancouver homeowners and renters, the calculus is indirect but real. Tech employment density in specific postal codes — V6A, V6B, V5T — has historically supported rental premium compression in those areas. When funded startup hiring slows, the premium that a two-bedroom in Mount Pleasant commands over a comparable unit in Burnaby narrows. The rental premium differential in those corridors has already softened from its 2022 peak, and a prolonged seed-funding drought in AI would extend that softening. This is not a collapse scenario. It is a slow pressure release.
Given the current BC assessment climate, charter banks are also adjusting — quietly. Small-business lending desks at major branches serving the Gastown and Yaletown tech corridors have tightened criteria for founder personal guarantees on commercial leases and operating lines. When the VC backstop for a startup's burn rate is uncertain, the bank's exposure on any co-signed instrument rises. That tightening does not make headlines, but it is a real constraint on working capital available to pre-revenue AI companies trying to bridge to their next institutional round.
The Contrarian Case for the Funding Drought
A seasoned LP who has watched three BC tech cycles would argue the hand-wringing about the pre-revenue funding gap is healthy market discipline, not ecosystem failure.
The AI seed frenzy of 2023 and early 2024 produced a cohort of Vancouver companies with impressive demo days and zero defensible data moats. Most of them quietly folded or pivoted by mid-2025. A source familiar with several of those cap tables, who asked not to be named, put it plainly: "The companies that couldn't raise in 2025 mostly couldn't raise because they hadn't built anything a customer would pay for. The market was right."
If the only capital writing $1M+ cheques to pre-revenue AI teams right now is hard to find, that is the market correctly pricing the asymmetric risk of foundation-model-dependent products with no proprietary training data and no enterprise contracts. Founders frustrated by this environment are often the ones who should spend another six months building, not the ones being failed by the ecosystem.
What Founders Should Actually Do in Q2 2026
The funding environment in BC right now rewards founders who understand their actual capital stack — not the one described in Slack rumours about funds that may not exist.
SR&ED is the starting point, not an afterthought. Hire a specialist, not a generalist accountant. The difference in recovered cash for a 10-person engineering team in Vancouver can exceed $500K annually when federal and provincial credits are properly stacked. That is non-dilutive runway that does not require a pitch deck.
The BCSC public registry is a free tool that every founder should use before spending pitch energy on any fund they cannot verify. If a fund is writing $1M+ cheques in BC, it is in that registry or operating under a documented exemption. If it is not findable there, the rumour is not a lead — it is noise.
The cross-border playbook is broken for now, but it is not permanently broken. US investor participation in Canadian pre-seed rounds dropped sharply in H1-2025. That drop creates a window for domestic angel syndicates and micro-funds to set terms they could not have set in 2022. Founders who close domestic rounds in this environment will have cleaner cap tables and more founder-friendly terms than the cohort that raised with US lead investors at the peak. That is a structural advantage, not a consolation prize.
The $24M in national AI seed investment in H1-2025 is a real constraint. It is also a real opportunity. In a market this thin, a credible pre-revenue AI team with a defensible data position and a real enterprise use case is not competing against hundreds of funded alternatives. The gap between the rumour and the registry is where the actual opportunity lives.
