Co-authored by articling student Alexander
Foulger-Fort.
As artificial intelligence, data centres and electrification drive demand for power at an unprecedented pace, Canada's energy sector is being pushed to adapt—fast.
On May 1, 2025, Gowling WLG was pleased to host its annual Canadian Energy Outlook at its offices in Toronto, Ontario, and brought together representatives from power generators, distributors, utilities, data centre developers, regulators, and members of the Gowling WLG team.
This year's discussion focused on the intersection of AI, energy, and data centres—exploring how this convergence is unfolding in Canada and the uncertainties they raise for the broader energy system. The discussions highlighted not only the technical and policy challenges ahead, but also the need for adaptability and collaboration to navigate a rapidly evolving landscape.
The observations and commentary provided below reflect only the authors' impressions of the discussions and should not be interpreted as representing the official views of any Energy Outlook participants.
Context and challenges
As AI, digital infrastructure, and electrification continue to expand, demand for electricity is expected to rise significantly. Yet despite the momentum, there remains substantial uncertainty about the scale, timing, and location of that growth. This uncertainty presents real challenges for electricity system planners, regulators, and infrastructure developers, especially where the pace of grid expansion must align with commercial demand that has yet to fully materialize.
The Energy Outlook discussions—particularly around data centres and AI-related infrastructure—highlighted the uncertainties surrounding demand forecasts, the constraints posed by interconnection and siting, the complications introduced by supply chain disruptions, the evolving requirements to maintain system reliability, and the role of governments in creating a policy environment which facilitates the evolution of the sector.
Taken together, these issues reveal a system in transition—where long-term investment decisions must be made in a context of imperfect information, and where coordination between market actors is essential to avoid costly mismatches between demand and capacity.
Data centres
Data centres are emerging as both a challenge and a strategic priority for the energy industry.
On one hand, their rapid growth and significant electricity requirements risk straining an already pressured grid, particularly at a time when capital investment decisions must be carefully justified. From a traditional utility planning perspective, data centres can be difficult to rationalize, as their long-term return on investment and net system value remain uncertain.
However, as the economy and public services become increasingly digitized, data centres are poised to become critical pieces of social infrastructure—supporting everything from healthcare and education to financial services and government operations. Recognizing this duality is essential: while the immediate benefits may be hard to quantify, ensuring reliable energy access for data infrastructure will be foundational to the functioning of future society.
Uncertain and evolving demand outlook
One of the key themes from the Energy Outlook this year was the emphasis on understanding the uncertainty surrounding future demand growth. While AI is widely expected to drive significant increases in power consumption, and current requirements may already be substantial, there remains considerable uncertainty about how this will evolve in the coming years.
Will AI systems improve their own efficiency, reducing their energy needs over time? Or, could those efficiency gains lead to broader adoption and ultimately even greater demand? These dynamics are still unfolding, and the long-term trajectory is far from certain.
This uncertainty is particularly relevant for those involved in the energy ecosystem such as generators, transmitters, and distributors, who must make long-term capital allocation decisions without clear visibility into future load. Their infrastructure planning depends not on broad projections, but on concrete signals from initial customers such as data center developers, hyperscalers, and large AI infrastructure firms.
However, these initial customers also rely on demand signals from businesses that require computing power before committing to major infrastructure investments. As a result, the entire system hinges on a chain of demand commitments, where each layer waits for greater certainty from the one before it. Until those commitments begin to solidify, the rest of the value chain remains in a holding pattern, balancing the risk of underbuilding against the cost of overcommitting to capacity that may not be needed.
Against this backdrop, governments and system planners are increasingly grappling with how to allocate limited grid capacity in a way that aligns with long-term policy objectives. Under current rules, non-discriminatory access ensures that any eligible customer, data centers included, can request and receive capacity if it is available.
However, this open-access approach may unintentionally crowd out future electrification priorities, such as electric vehicle charging or industrial process decarbonization, especially if those users come later and find that capacity has already been committed. One speaker noted that when demand from data centers began rising, policy focus remained heavily tilted toward the automotive and industrial sectors, contributing to a mismatch between connection policy and emerging needs.
To manage these tensions, system operators are turning to scenario-based planning and predictive analytics to better anticipate future connections and assess their infrastructure impacts. Regional planning cycles now play a larger role in evaluating cumulative system demands, while some stakeholders called for more differentiated approaches to valuing and allocating capacity. Not all capacity, it was argued, is created equal—its location, reliability requirements, and opportunity cost should inform how risk is assigned.
Without targeted frameworks, utilities face regulatory and financial exposure if they overbuild based on speculative or shifting forecasts. As Ontario's electricity system navigates competing demands, and economic, environmental, and political uncertainty—more nuanced, forward-looking planning will be essential to ensuring both flexibility and fairness in grid access.
Governments also play a critical role in shaping a stable and forward-looking policy environment to support the growth of Canada's electricity sector, particularly as it faces mounting pressure from rising energy demand and investment uncertainty. Clear, consistent, and coordinated policy signals—such as long-term decarbonization targets, regulatory reform, and investment incentives—are essential to guide private and public sector actors in making timely infrastructure and generation decisions.
By de-risking investment through supportive frameworks and fostering collaboration across jurisdictions, governments can help create certainty to ensure that the sector expands reliably, affordably, and sustainably to meet future energy needs.
Supply chain risks
As large energy users like data centers seek grid connections, supply chain constraints have become a significant challenge alongside issues like siting and transmission access. Speakers noted that even with strong demand signals, long lead times for key equipment—especially transformers and turbines—can delay projects by years.
Some utilities and developers have responded by pre-ordering infrastructure on a speculative basis, taking calculated risks to attract business. This strategy worked during periods of high activity, such as the EV boom, but is difficult to sustain, especially when rising costs are passed through to consumers.
While supply chain delays are global in nature, well-capitalized firms may mitigate risks by placing early orders and reallocating equipment across projects. Smaller players, however, lack that flexibility. In some cases, developers adapt their system designs around what equipment is available, but that requires significant creativity and risk tolerance.
Despite these challenges, some argued that supply chain issues are not the primary barrier. The larger concern is the lack of a market structure and regulatory signals that enable proactive investment. Even with capital available, firms are hesitant to take procurement risks without a supportive regulatory or commercial framework.
Geopolitical constraints further narrow sourcing options, compounding the problem. Still, many believe that with better planning, clearer signals, and more adaptable design, supply chain risks can be managed. But without these conditions, delays and higher costs will continue to affect infrastructure deployment.
System reliability concerns
When considering what Ontario and federal governments must do to support data center development, one speaker emphasized that, in the near term, progress will depend on close coordination on a case-by-case basis.
While system operators have the authority to reject projects, the preference is to enable connections, provided certain conditions are met. These include locating in areas with available transmission capacity, coordinating with the Independent Electricity System Operator on operational issues like ramping, participating in restoration efforts during outages, and meeting increasingly critical cybersecurity standards. Large-scale data centers introduce new system stability risks. As such, a cyber event causing a sudden ramp-up, drop, or outage could pose grid-wide reliability risks.
These emerging realities mean that integration will require not only technical adaptation but also new forms of collaboration between developers, grid operators, and policymakers.
Interconnection challenges
Building on the discussion of uncertain demand and the evolving nature of AI-driven load growth, another key factor that shapes infrastructure planning is the availability and flexibility of the transmission system.
Access to transmission capacity is not only a prerequisite for connecting new load, but also a critical constraint in determining where data centers and AI infrastructure can be sited. Additionally, proximity to urban centres—where data is generated and consumed—is often essential for data centres to reduce latency, further narrowing the range of viable locations. Projects must either locate where transmission is readily available or be flexible enough to adapt to the limitations of the existing grid. In many cases, that flexibility becomes the determining factor in whether a project can move forward.
Despite a surge in expressions of interest, in one instance, a speaker shared that out of 6,000 megawatts of potential data centre load, only a third was expected to materialize. So, a significant portion of this interest remains speculative, and only a fraction advances to the point of securing interconnection. In another instance, approximately 1,500 megawatts are in the process of connecting, with expressions of interests several times higher. But even among the 'committed' projects, some developers are considering deferrals.
This underscores a core challenge: the timing and firmness of customer commitments remain fluid, which complicates transmission planning and capacity allocation. Until these commitments solidify, system operators must plan around a moving target, balancing readiness with the risk of stranded investment.
As one speaker put it "[developers] don't proactively build data centres, and we don't proactively build transmission lines to accommodate data centers." The process is inherently reactive: developers wait for customer demand before building data centres, and system operators wait for firm developer commitments before investing in transmission infrastructure.
Transmission constraints
The issue of transmission constraints was a key topic of discussion. It was noted that some of the most pressing infrastructure challenges are not at the last mile or generation level, but upstream within the core transmission system itself. These are deep-rooted structural issues relating to how the main transmission feeds were historically built and interconnected that now limit the ability to deliver power into high-growth areas.
One speaker noted that their project's available capacity was cut by 75 per cent due to such systemic constraints. Addressing these transmission bottlenecks will require substantial capital investment, the strategic use of emerging technologies to improve coordination between supply and demand, and clear provincial direction to focus on resolving foundational system limitations.
At the same time, the discussion turned to the tension between near-term connection requests—particularly from large data centers and longer-term electrification goals. Under current non-discriminatory access policies, data centers are being connected as they apply. But this raises the risk that by the time broader electrification initiatives take shape, such as EV charging networks or industrial process decarbonization, there may be no remaining capacity to accommodate smaller but socially critical loads.
For example, if a small EV charging operator needs just ten megawatts, that capacity might no longer be available. As one participant asked, how do we reconcile immediate, high-volume connections with the need to preserve grid headroom for the energy transition? Without careful planning, today's connection policy could constrain tomorrow's electrification.
Looking forward: Collaboration and adaptability
Despite the complexity and uncertainty discussed throughout the Energy Outlook, a recurring message was one of opportunity, but only if market participants are willing to be flexible, transparent, and proactive. The discussions emphasized the importance of early collaboration between generators, utilities, developers, policymakers, and customers to navigate uncertainty and align expectations. Tools like scenario-based planning, regional coordination, and predictive analytics are already being used to better anticipate who is coming and when.
There was also a clear call for openness to new technologies, new market participants, and new ways of allocating and valuing capacity. While energy systems cannot be built speculatively, it also cannot afford to be entirely reactive. A more adaptive planning framework, coupled with clear signals and shared accountability, will be essential.
As one speaker put it, the key is "keeping the door open"—engaging early, planning around reasonably predicted scenarios, and being prepared to pivot as conditions evolve.
The power sector in Ontario, and across Canada more broadly, may be entering a period of historic transformation. With coordination, adaptability, and sustained commitment from all market participants, it has the potential to meet both present and future challenges.
Read the original article on GowlingWLG.com
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
