You are receiving the Community Edition of today’s newsletter. Become a member to receive exclusive perks and event invites.

In 2024, American regulators caught a Toronto investment firm lying about the one thing its entire business was built on.

Delphia claimed for years that its proprietary AI, fed by client data, could predict breakout companies. But that AI never existed. The SEC investigated and found the firm had no such capability, fining Delphia's U.S.-registered advisory arm $225,000 USD in 2024 for false and misleading claims.

That is AI washing in its purest form: making a bold claim with nothing to back it up.

It's also precisely the kind of cautionary tale that brought an end to easy AI pitches. Between 2023 and 2024, slapping "AI-powered" onto a pitch deck could instantly boost a startup's valuation—long before anyone audited the actual tech stack. That valuation premium has dissolved, driven out by investors demanding real proof.

Founders haven't stopped leaning into the buzzword. If anything, more pitch decks feature "AI-powered" today than two years ago. What has changed is how skeptically investors view that label on its face—and how thoroughly they scrutinize what lies behind it.

Two years ago, asking which AI model you were using passed for diligence. That question yields little insight today, prompting more rigorous funds to establish formal frameworks for verifying technical claims.

When Deloitte Canada published its AI due diligence framework in June 2026, it boiled the evaluation down to four essential questions:

Venture firm CRV frames the third question more directly: what happens to your business the day OpenAI or Anthropic builds your exact feature and offers it for free? Consider how you would answer that.

If proprietary customer data lives in your system and switching platforms requires rebuilding entire workflows, you have built defensible value. But if losing a feature to a base-model update wipes out most of your user base, the product was never truly yours. It belonged to the underlying model.

None of Deloitte's four criteria require building a proprietary foundation model from scratch. Successful companies are regularly built on top of third-party platforms like GPT or Claude. What separates enduring products from fleeting wrappers comes down to three factors:

If you can clearly demonstrate at least one of these advantages, you have a viable pitch. If not, it is far better to address that gap before entering the boardroom.

While some founders assume AI washing is primarily a concern for those fundraising in Silicon Valley, Canadian regulators are scrutinizing technical claims with equal rigour.

The Canadian Securities Administrators (CSA), including the Ontario Securities Commission (OSC), explicitly targets AI washing—defining it as false or exaggerated AI claims designed to mislead investors. Notably, when the OSC audited Canadian investment funds in November 2025, those funds passed clean.

Capital moves across borders seamlessly. Toronto-based Radical Ventures, global AI-focused VC firm, backs Cohere and Waabi while deploying capital globally, while U.S.-based Forum Ventures chose Toronto to establish its AI venture studio. Whether incorporated in Canada or the U.S., founders building for a global market pitch to the same investor pools and face the same rigorous standards.

State clearly what you built, what you license from third-party models, and how your business model holds up if provider terms change tomorrow. That level of transparency carries far more weight in pitch meetings than the term "AI" ever did on its own.

Founders who answer these questions directly navigate due diligence efficiently; those who cannot end up spending weeks defending fragile assumptions.

To leave you with one final thought: if an investor asked today what happens when your core feature gets built into a foundation model for free, could you give a compelling answer in a single sentence?

Until next time,

Alex

Building something and want 70,000+ TechTO readers to see it? Reply to this email or reach out to us at [email protected] and tell us what you're working on. One founder gets featured every week.

This week’s feature is Kevin Callahan, who is also one of our speakers at Together Toronto on Sept 14. If you want to ask him about any of this in person, that's the place.

What he's building: One simple connection that lets an app talk to over 300 different blockchains, instead of needing a separate setup for each one.

What problem is he solving: Right now, teams have to build and fix their own connector for each chain they use. Uniblock solves the “too many chains, too many providers, too much downtime” problem by giving developers a single, reliable, smartly routed API that works across the whole multi-chain ecosystem.

One learning: Kevin left his tech career to build Uniblock and one of the reasons why it keeps on growing is because he only works with S-tier people. “Startups without top talent are dead on arrival. Uniblock is the strongest team I have ever worked with. The output relative to our resources is extraordinary.”

If you have been meaning to meet potential collaborators, compare notes with other founders, find your next hire, or simply get out from behind the screen and be part of Toronto’s most connected tech community, this is the room.

Hiring? Currently looking for your next gig? Check out the 1000+ jobs on our job board or post yours.

Want to showcase your company, events, and opportunities to thousands of tech leaders, professionals, and investors across the country?

Reach out to us for partnership and advertising opportunities.

The best answers come from founders who've already solved the problem.

So we're opening the floor. What's one challenge you're facing right now where you could use a second opinion?