Power Corp Trio Backs Sagard's $500M AI Fund in Rare Joint Move

Power Corporation of Canada and two of its publicly traded portfolio companies—Great-West Lifeco and IGM Financial—announced Tuesday they've made a joint anchor investment in Sagard's newly raised AI-focused private equity fund, marking one of the rare instances all three entities have coordinated capital deployment behind a single strategy.

The move puts hundreds of millions from Canada's financial establishment behind Sagard's thesis that the next wave of enterprise software value creation sits in verticalized AI applications rather than horizontal infrastructure plays. Sagard, the alternative investment arm controlled by Power Corp, closed the fund at approximately $500 million, according to sources familiar with the raise who weren't authorized to disclose the final figure publicly.

What makes the announcement unusual isn't the size—mid-market AI funds have raised comparable sums over the past 18 months—but the coordinated backing from Power Corp's three largest operating entities. Great-West and IGM typically pursue independent investment strategies despite shared ownership. Their joint participation with the parent company signals institutional conviction that AI specialization represents a structural shift worth concentrating exposure around.

"We're not building another generalist tech fund with an AI tilt," a Sagard partner said in the announcement, though the firm declined interview requests. "This is purpose-built capital for companies applying frontier models to specific industry workflows where incumbents can't move fast enough." The fund targets enterprise software companies generating $10-50 million in revenue that are embedding large language models, computer vision, or predictive analytics into vertical solutions for healthcare, financial services, supply chain, and legal sectors.

Power Corporation owns roughly 67% of Great-West Lifeco and 62% of IGM Financial, but the three entities maintain separate investment committees, regulatory obligations, and fiduciary duties to distinct shareholder bases. Great-West operates as a multinational life insurance and asset management company with $2.3 trillion in assets under administration. IGM runs wealth and asset management businesses through IG Wealth Management and Mackenzie Investments.

Public market investors in Great-West and IGM have historically pushed back when the companies appear to function as extensions of Power Corp's strategy rather than maximizing standalone returns. Joint investments require each entity to independently justify the allocation to its own board and public shareholders, creating friction that usually leads to separate, staggered capital deployment even when interests align.

The last notable instance of coordinated investment across all three came in 2019 when they collectively backed Wealthsimple's Series D round—and even then, the checks came in at different valuations across multiple tranches. Before that, you'd have to go back to the early 2000s infrastructure plays in China to find parallel deployment at this scale.

That this fund cleared all three approval processes simultaneously suggests the investment thesis resonated not just as an alternative asset diversification play, but as a strategic position on where enterprise value creation is heading. For Great-West, AI tools that improve underwriting, claims processing, and actuarial modeling represent direct operational leverage. For IGM, verticalized wealth tech—robo-advisory 2.0, if you will—could reshape the advisory businesses it operates.

Sagard's fund targets what it calls "applied AI"—companies that aren't building foundational models but are instead wrapping OpenAI, Anthropic, or Google's models into software that solves narrow, high-value problems in specific industries. Think AI-native legal research platforms, radiology workflow tools that pre-screen imaging studies, or supply chain optimization software that predicts component shortages three quarters out.

The strategy reflects a broader shift in institutional LP appetite. After 2023's generative AI hype cycle saw capital flood into infrastructure and horizontal tooling, limited partners are now asking which application-layer companies will actually generate durable revenue and margin expansion. Vertical AI plays—where the software is built for a specific buyer with embedded industry expertise—are emerging as the answer.

"The horizontal plays got too crowded too fast," said one institutional allocator who reviewed Sagard's fund materials but declined to participate. "Everyone's building the same developer tools, the same internal chatbots, the same workflow automation. Sagard's betting that defensibility lives in knowing healthcare reimbursement codes or supply chain logistics better than a generalist ever could."

The fund's deal flow focuses on growth-stage companies—past product-market fit but pre-scale—where Sagard can take board seats and provide operational support around go-to-market, compliance navigation, and enterprise sales cycles. Check sizes range from $15 million to $50 million, targeting ownership stakes between 15-30%. The firm plans to make 10-12 core investments over a three-year deployment period.

Sagard hasn't disclosed whether the fund operates as a traditional committed capital vehicle or an evergreen structure, though sources suggest it follows a standard 10-year fund life with typical 2-and-20 economics and a preferred return hurdle before carry kicks in. The Power Corp family's anchor commitment likely represents 30-40% of total capital, with the remainder coming from external institutional LPs including university endowments and pension funds.

Healthcare represents the fund's largest sector allocation. Sagard sees opportunities in AI-driven clinical documentation, prior authorization automation, and revenue cycle management tools that reduce administrative burden for health systems. One person familiar with the fund's pipeline said Sagard has already signed two term sheets—one with a radiology workflow company and another with a platform automating Medicare Advantage plan comparisons for brokers.

The institutional shift toward vertical AI funds reflects broader disillusionment with the horizontal infrastructure wave that dominated 2022-2024 fundraising. Vector databases, AI ops platforms, and fine-tuning tools all promised to be critical infrastructure for the AI economy. Most delivered modest returns and fierce competition.

Limited partners are now asking a sharper question: where does proprietary value actually accumulate in the AI stack? The answer increasingly points to application-layer companies with deep vertical expertise, customer lock-in through workflow integration, and margin structures that improve as models get cheaper and better.

"We stopped underwriting infrastructure deals six months ago," said a senior investment director at a $40 billion public pension fund. "The model providers are commoditizing everything below them. The only durable businesses are the ones selling to end users who don't care what's under the hood—they just want the workflow solved."

This thesis explains why Sagard structured the fund around sector verticals rather than horizontal capabilities. A legal AI company competing with Casetext or Harvey isn't selling "AI-powered search"—it's selling faster brief writing, better contract review, or more accurate case law prediction. The buyer doesn't evaluate it against other AI companies; they evaluate it against their current workflow.

That verticalization creates defensibility. A supply chain optimization tool built for automotive manufacturers can't easily pivot to consumer packaged goods. The data pipelines, integrations, and domain expertise are specific. Competitors can't replicate it by swapping in a better foundation model.

Power Corporation's willingness to commit capital across its three largest entities suggests the firm views vertical AI not as a thematic bet but as a structural realignment of enterprise software. The family office-style holding company has historically concentrated capital in financial services, insurance, and wealth management—sectors where AI-driven automation represents existential change rather than incremental improvement.

Great-West's participation is particularly telling. Life insurers have spent decades building actuarial models, underwriting systems, and claims platforms that rely on historical data and rules-based logic. AI tools that can predict policyholder behavior, automate underwriting decisions, or flag fraudulent claims before payout represent operational leverage that flows straight to the bottom line. Great-West isn't investing in Sagard's fund as portfolio diversification—it's buying a window into the technology that will reshape its core business.

Sagard operates multiple strategies including private equity, venture capital, real estate, and royalty financing. The firm manages approximately $6 billion in assets across its platforms, though it doesn't break out performance by fund. Its private equity arm has historically focused on North American middle-market companies in healthcare, business services, and industrials—sectors where operational improvement drives returns more than multiple expansion.

The AI fund represents a departure from that profile. It's sector-agnostic in the sense that it will invest across healthcare, financial services, legal, and supply chain—but it's hyper-focused on a single technology thesis. That suggests Sagard sees vertical AI as a category unto itself, not a subsector of software or a feature of existing enterprise platforms.

The firm's prior venture investments include stakes in Dialogue (virtual healthcare), Nuvei (payments infrastructure), and Well Health Technologies—all companies operating in regulated sectors where technology adoption lags consumer markets. That pattern holds with the AI fund: Sagard isn't chasing horizontal dev tools or consumer AI apps. It's targeting industries where incumbents move slowly and buyers pay premium prices for workflow solutions.

"Sagard doesn't do moonshots," said a founder who pitched the firm in 2024 but ultimately raised from a different investor. "They want to see revenue, they want to see customer retention, and they want to see a path to profitability that doesn't depend on raising another round. If you're pre-revenue with a research lab and a demo, you're not their company."

University endowments and pension funds have struggled to get vertical AI exposure without backing generalist tech funds that happen to own one or two relevant companies. Specialized funds like Sagard's offer concentrated exposure to the thesis LPs actually want—application-layer businesses with revenue and margin defensibility—without the dilution of a broader software portfolio.

That positioning explains why Sagard likely had an easier fundraising path than horizontal AI infrastructure funds raised over the same period. LPs are underweight specialized vertical strategies and overweight generalist funds that all look similar on paper. A fund that says "we only back vertical AI in regulated industries" solves a portfolio construction problem for allocators trying to build non-correlated exposure.

The companies Sagard targets aren't trying to be OpenAI or Anthropic. They're wrapping frontier models into software that solves narrow problems buyers will pay for immediately. A healthcare revenue cycle management company using AI to predict claim denials isn't building a model from scratch—it's fine-tuning GPT-4 on millions of historical claims, then embedding that intelligence into workflow software that billing departments already use.

That approach has structural advantages. Development costs are lower because the company isn't training foundational models. Time to market is faster because the product integrates with existing systems rather than replacing them. Customer acquisition is easier because buyers evaluate ROI against current workflows, not against theoretical future states.

The risk is commoditization. If the underlying models improve fast enough—and if integration becomes standardized enough—then vertical AI companies could lose defensibility. A radiology workflow tool that's differentiated today because it uses Claude 3.5 might become undifferentiated if every competitor can swap in Claude 4 next year.

Sagard's bet is that workflow integration, proprietary data, and customer relationships create enough friction to sustain margins even as the underlying models improve. Whether that holds depends on how fast incumbents move—and whether they build, buy, or partner their way into vertical AI capabilities.

"The real question isn't whether vertical AI companies can build good products," said a managing director at a competing growth equity firm. "It's whether they can build them faster than Epic Systems or Salesforce or Workday decides to embed the same capabilities into platforms that already have distribution and customer lock-in. That's the race."

Sagard's announcement leaves several questions unresolved. The firm didn't disclose the fund's final size, though $500 million appears consistent with comparable vertical AI strategies raised over the past year. It didn't name external LPs beyond the Power Corp family, though sources suggest the fund includes at least two large public pension plans and one university endowment.

The firm also didn't clarify whether it will co-invest with other Sagard funds or maintain strict separation between the AI vehicle and its broader private equity strategy. That matters for portfolio companies: if Sagard can pull operational resources from its healthcare or business services teams, that adds value beyond capital. If the AI fund operates in isolation, it's just another check.

Most notably, Sagard didn't specify how it will handle the build-versus-buy decisions that every vertical AI company eventually faces. When a portfolio company hits $50 million in revenue and Oracle or Microsoft makes an acquisition offer, does Sagard take the exit or push for continued independence? The answer shapes returns—and reveals whether the firm believes vertical AI businesses can scale to standalone public market valuations or if they're ultimately features that get absorbed into platforms.

The fund's structure—whether it includes continuation vehicle optionality, secondary sale provisions, or long-term hold flexibility—also remains undisclosed. Those mechanics matter in a market where AI companies are scaling faster than traditional software businesses but may require longer hold periods to reach exit valuations that justify the risk.

Sagard's raise is the latest in a wave of specialized AI funds launched over the past 18 months as institutional LPs seek concentrated exposure beyond generalist tech portfolios. Felicis Ventures raised a $900 million AI-focused vehicle in late 2025. Bessemer Venture Partners closed a $3.3 billion fund with explicit AI verticalization mandates. Greylock shifted its entire strategy toward application-layer AI after missing the infrastructure wave.

The trend reflects a broader market realization: AI infrastructure returns will accrue to a handful of model providers and cloud platforms, while application-layer returns will distribute across hundreds of vertical-specific companies. For LPs looking to capture AI upside without betting on the next OpenAI, vertical strategies offer a more diversified path.

That doesn't mean vertical AI funds are guaranteed winners. The category is crowded, the exit landscape is uncertain, and the risk of incumbent platform absorption remains real. But for institutional allocators who spent 2023-2024 chasing horizontal AI infrastructure and came away with modest returns, vertical specialization represents a reset—and a chance to get positioning right before the application layer consolidates.

Whether Sagard's fund delivers outperformance will depend on deal selection, operational value-add, and exit timing. But the coordinated backing from Power Corp's three largest arms suggests the firm views this not as a tactical allocation but as a strategic position on the next decade of enterprise software. That's the real signal in Tuesday's announcement—not the dollar figure, but the conviction behind it.