Alex McIsaac here with my 10 predictions for 2026 on how AI, technology, capital, and business models will evolve and the trends shaping where we allocate capital at Northside Ventures.

1. Agent-to-agent systems go mainstream

AI agents evolve from copilots into autonomous systems that coordinate directly with other agents across complex workflows. Long-running, multi-step processes in finance, business development, compliance, security, procurement, and operations increasingly run with humans removed from execution and retained primarily for oversight. As autonomy rises, demand grows for observability and orchestration layers, identity and permissions, auditability, simulation environments, and deterministic shutdown mechanisms, turning agent infrastructure into a core software category rather than an edge feature.

2. Enterprise AI moves from pilots to production

Agents will be embedded directly into core, revenue-impacting enterprise workflows rather than confined to paid pilots and internal productivity experiments. Buyers increasingly will price agents against the fully loaded cost of human labor and not SaaS with a pay per seat model. As agents prove more consistent and reliable than humans on repeatable work, enterprises begin restructuring teams where human roles are rendered economically redundant.

3. AI infrastructure spend hits new highs amid regional friction

Global AI infrastructure investment will exceed $600B annually across data centers, networking, silicon, power, and cooling as demand continues to outstrip supply. Capacity constraints drive tighter coordination between hyperscalers, chipmakers, utilities, energy providers, and municipal / state / provincial governments, while pushing overflow workloads toward neo-clouds and specialized operators. In parallel, a new layer of infrastructure service companies will emerge to support rapid data-center deployment at scale, spanning power generation, modular construction, cooling, orchestration, and financial operations.

4. Physical and industrial AI capture an outsized share of venture capital

As software creation becomes increasingly commoditized by AI tools, venture capital continues to concentrate around categories tied to the physical world, where defensibility, capital intensity, and real-world constraints create durable moats. Robotics, industrial automation, aerospace, mining, and defense benefit from AI-driven advances in perception, reasoning, and autonomy that unlock new use cases across complex, unstructured environments. This momentum is reinforced by deglobalization and reshoring efforts, as governments and enterprises invest in domestic manufacturing capacity and cost reduction through industrial robotics, drawing sustained follow-on capital into these sectors.

5. Canada's defense and security sectors boom

Canada’s defense and security sector enters a sustained expansion as government spending resets after decades of underinvestment. Defense outlays are accelerating toward NATO’s 2% of GDP target — implying annual spend rising from roughly $40–45B today to $75–80B with some discussions pointing toward 5% by 2035, or $180B+ annually at current GDP levels. This capital increasingly flows into domestic production, modernization, and AI-enabled capabilities across naval, air, land, cyber, drone and counter-drone technologies and space systems, drawing private investment into defense-adjacent AI, robotics, and advanced manufacturing.

6. Stablecoins cement their place in the global payment system

2025 marked the inflection point, driven by regulatory clarity, public-market validation, and rapid adoption by fintechs, payment providers, and financial institutions, pushing stablecoins into credible, regulated infrastructure. In 2026, the focus shifts from growth to entrenchment as stablecoins become a default rail for cross-border B2B payments, treasury movement, and emerging-market remittances, where speed, cost, and settlement finality materially outperform legacy systems. By year-end, stablecoins are no longer viewed as crypto products but as a core component of global payments infrastructure, with competition centered on compliance, distribution, and embedded use cases.

7. AI valuations stay extreme while non-AI capital tightens

Elite AI teams continue raising $20–30M Series A rounds at hundreds of millions in valuation, often ahead of scaled commercial traction, while top growth-stage AI companies secure $100M+ rounds at multi-billion-dollar valuations. Capital increasingly concentrates around perceived category leaders as investors prioritize platform leverage, data moats, and model-driven defensibility. In contrast, non-AI companies face prolonged fundraising pressure as capital selectivity increases, amplified by record-sized funds that must deploy large checks into fewer, conviction-driven outcomes.

8. 2026 becomes the strongest liquidity year since 2021

2026 delivers the second-best tech liquidity year of the past decade, behind only 2021, as expectations reset following uneven IPO outcomes in 2025. Public markets reopen selectively for higher-quality, AI-exposed companies, while private markets see increased secondary volume and targeted M&A as incumbents acquire capabilities faster than they can build internally. At the same time, timelines between rounds lengthen as milestone expectations rise and companies operate more capital-efficiently.

9. Gen-Z and university dropout founders surge

A record number of Gen-Z founders leave university to start AI-native companies, emerging directly from campus ecosystems, hacker houses, and tightly networked builder communities. As execution tooling improves, the value of university increasingly shifts toward admissions signaling, early networks, and community formation rather than completion. Access to powerful models, tooling, and agents compresses the gap between idea and execution, enabling young teams to ship production-grade products with a level of speed and sophistication that historically required years of industry experience.

10. Organizations are structurally rethought and rebuilt around the agent economy

Newly formed companies will redesign org structures from AI-first principles, treating agents as team members across every function. The model shifts from individual productivity tools to coordinated teams of agents replacing entire human functions, delivering 30–40× productivity gains rather than incremental 30–40% improvements. Established companies face a classic innovator’s dilemma, forced to choose between launching AI-native subsidiaries that intentionally cannibalize the core business or implementing organization-wide AI change programs to remain competitive.