Artificial intelligence (AI) may be having its gold-rush moment, but somebody still has to know where all the gold is buried.
For businesses, that means data.
Companies can buy faster chips, build enormous data centres and deploy shiny new AI agents. Yet those systems aren’t terribly useful if a company’s information is scattered across decades of documents, customer records and software systems that can’t safely talk to one another.
That rather unglamorous problem kept showing up beneath the much flashier discussions at Canada’s Investment Summit.
One main-stage panel was literally titled “Invented Here, Scaled Here: Financing Canada’s Frontier Technology Champions.” Prime Minister Mark Carney also said Canada wants to capture value across the “entire intelligence infrastructure stack,” stretching from electricity and compute to cloud services and frontier AI.
That brings me to a Canadian technology company investors may have forgotten: OpenText (TSX: OTEX).
Source: Getty Images
The summit connection
OpenText wasn’t presenting on the main stage. However, Thomas Jenkins, chair of both OpenText’s board and OpenText Ventures, appeared on the summit’s official participant list.
It’s clear why the company was at least present. The summit repeatedly returned to one Canadian problem. We invent good technology, then struggle to scale it.
OpenText is what happens when Canada actually succeeds. The Waterloo-born company now serves organizations globally with software that manages, secures, and organizes enterprise information. In an AI world, that information becomes increasingly valuable because AI models need reliable corporate data before they can produce reliable answers.
Investors looking through Canadian AI stocks therefore shouldn’t assume all the money will end up with chip manufacturers and data centre builders. The software sitting on top of that infrastructure could be just as interesting.
OpenText just made the summit thesis real
Timing helps enormously here. On September 16, OpenText announced a strategic partnership with another company featured prominently at the summit: Cohere.
The two plan to combine Cohere’s agentic AI technology with OpenText’s enterprise-data platform for governments and heavily regulated industries such as financial services and healthcare.
That’s a clever marriage. Cohere provides AI models and agents. OpenText provides the governed, secure corporate information those agents need to actually do useful work without treating confidential records like suggestions.
Chief executive officer Ayman Antoun has described OpenText as the “secure data foundation in the AI stack.” Now the numbers are beginning to support that pitch.
A reasonable price
During earnings, fiscal 2026 cloud revenue increased 5.5% to US$1.96 billion, while enterprise cloud bookings surged 22.5% to US$947 million. Even better, OpenText generated US$808 million in free cash flow, up 17.5%, while adjusted earnings before interest, taxes, depreciation, and amortization (EBITDA) reached US$1.9 billion.
That’s considerably different from buying an AI company that comes equipped with breathtaking revenue growth and an equally breathtaking absence of profit. The stock recently traded around $32.75 on the TSX, approximately 42% below its $56 52-week high. The U.S.-listed shares trade for roughly six times forward earnings.
For investors considering Canadian technology stocks, that’s an unusual combination of AI exposure, cash flow and a beaten-down valuation.
Foolish takeaway
Growth remains OpenText’s biggest problem. Fiscal 2026 total revenue increased just 1.5%, while annual recurring revenue rose only 1.3%. The company also carried about US$5.7 billion of long-term debt at year-end.
If AI fails to accelerate organic growth, a low valuation can stay low for a remarkably long time. Yet OpenText paid down roughly US$649 million of debt during fiscal 2026 while returning US$677 million through dividends and share repurchases.
That’s why I like the setup. Canada’s summit spent two days asking how the country can turn its technology expertise into global businesses. OpenText already did that once.
If its enormous reservoir of enterprise data becomes the plumbing underneath corporate AI, it might be preparing to do it again.
