13 February 2026
BlackRock has advised clients to move from viewing artificial intelligence (AI) purely as software to viewing it as an energy issue. In its Global Outlook 2026, the BlackRock Investment Institute argues that AI development is reaching physical limits and that electricity is the undervalued limiting factor for investors.
A startling warning from the report is that AI-powered data centers could consume up to 24% of US electricity by 2030. This would have far-reaching implications for everything from utility investments to the siting of industrial facilities.
This prediction raises crucial questions for the crypto sector: if access to energy becomes so scarce, what will that mean for the industry that has built its business model around converting cheap, intermittent power into Bitcoin?
In 2025, narratives emerged about the potential synergy between crypto and AI, stemming from the theory that AI agents would want to use crypto for payments instead of traditional finance. However, a competition for electricity could undermine this relationship in the future.
The mining sector has been embroiled in a political debate about energy waste for years. The sector's counterargument has always been operational: miners can act as a flexible load; they switch off when the grid is under pressure and absorb surplus energy when prices collapse.
In Texas, the Electric Reliability Council of Texas (ERCOT) has designed explicit programs for “large flexible customers, such as Bitcoin mining facilities,” and encourages limiting consumption during peak demand.
However, AI data centers come with different consumption patterns, different contract terms, and a different level of political support. They never want to shut down. They want constant power.
BlackRock's broader point is that the AI boom is unusually capital-intensive. The firm estimates total capital investment for AI development to be between $5 trillion and $8 trillion by 2030, with significant investments in computing power, data centers, and energy infrastructure.
What started as a race for chips has quickly evolved into a race for megawatts.
There's broad agreement that data center electricity demand is increasing rapidly, even as analysts question the ceiling. A Department of Energy announcement related to a report from the Lawrence Berkeley National Laboratory states that data center load growth in the US has tripled in the past decade.
Moreover, this demand is expected to double or triple by 2028. EPRI modeling from 2024, cited by Utility Dive, places U.S. data centers at 4,6% to 9,1% of U.S. electricity generation by 2030, depending on AI adoption and efficiency gains.
An explanation from the World Resources Institute, citing a Berkeley Lab study, points to a consumption of 6,7% to 12% of US electricity by 2030. BlackRock's "up to 25%" framing is on the aggressive end of that spectrum and is intended to be provocative. But even the lower-end scenarios would be enough to inflame energy markets and harden the politics surrounding the grid over who gets first access.
Reuters reported that utilities and grid operators are already adjusting their rate structures and rules as hyperscalers and colocation firms fight for capacity, especially in hotspots like Texas and Northern Virginia.
This is the context in which Bitcoin miners now find themselves. They are large, mobile energy users and are first in line in areas with abundant energy production or attractive prices. Until now, these characteristics seemed to be an advantage.
Bitcoin mining is incredibly simple on a physical level. Specialized computers perform hashing to secure the network, and electricity is the dominant cost. When power is cheap relative to the Bitcoin price and network difficulty, miners make a profit. When electricity is expensive, they shut down, relocate their operations, or go bankrupt.
This operational flexibility has become the sector's strongest argument as public scrutiny has increased. The U.S. Department of Energy estimates that crypto mining will likely account for 0,6% to 2,3% of U.S. electricity consumption by 2024. A small percentage in absolute terms, but large enough to be noticed in local politics and grid planning.
Texas is the cleanest case study, as the state's competitive energy market converts this flexibility into revenue. In a 2023 SEC filing, Riot Platforms reported that in August 2023, it had curtailed power consumption by more than 95% during peak demand, opting to forgo mining revenues to support ERCOT's reliability.
CryptoSlate ERCOT reported that a miner had paid $31,7 million in energy credits that month to shut down during a heat wave, a detail that captures both the value of flexibility and the speed at which the political situation can deteriorate.
When we compare this model to AI, training and operating large models requires constant energy and strict uptime. A hyperscaler signing a long-term contract demands reliable delivery, not voluntary scaling back.
If miners are the shock absorber, then AI is the shock creator.
And BlackRock's annual outlook effectively states that the shock is coming and there's nothing you can do about it.
In the mining handbook, "cheap power" often means stranded hydropower, excess wind power at night, or a friendly industrial tariff. But as data centers scale up, cheap energy becomes a moving target, as grid access itself becomes the bottleneck.
Interconnection queues and transmission delays are the new friction. Even if a region has energy production, the lack of infrastructure and approvals to transport it to a new 500-megawatt campus can be problematic.
NERC has warned of reliability threats due to the rapid growth in demand that coincides with AI, data centers, electric vehicles, and electrification, while traditional generators are being phased out and slow build-up processes persist. This is particularly relevant for miners, as their advantage is speed.
They can place containers at a location, power them up, and start mining faster than a conventional industrial plant can. But if the limiting factor becomes substation capacity and interconnection approvals, this speed turns into a regulatory battleground.
When energy markets tighten, legislators start looking for villains. Mining is often convenient because it feels optional, even to those unfamiliar with it. AI, on the other hand, is now being presented as crucial to national competitiveness, both to the public and to legislators.
This asymmetry will shape policy. It's easier to impose reporting requirements or additional fees on miners than on data centers that the local chamber of commerce is trying to lure in. It's also easier to frame mining as a speculative luxury and AI as the backbone of defense, productivity, and medicine.
If BlackRock is right that AI's energy footprint will become a macro risk, the political coalition supporting grid investments could broaden, but the pressure to prioritize "productive" burdens could also increase.
Miners could respond by pushing the narrative of flexibility harder. A Duke University report cited by Utility Dive states that the current US grid can handle significant new loads, provided it can be mitigated during periods of stress, and mining is capable of doing so. Many AI workloads, especially those for consumer products, typically cannot.
This creates a potential point of contention: miners as a controllable tax that helps integrate renewable energy sources, versus data centers as an inflexible tax. This argument has already surfaced in policy circles and hearings of the electricity commissions.
However, whether it wins depends on local economies and lobbying, not on internet debates.
Another route to adaptation already exists: switching from hashing to hosting.
The logic is simple. If you already own land, energy rights, and a substation, you have exactly what AI developers need most. And if your traditional business is volatile, the prospect of long-term cash flows from compute hosting is tempting.
CryptoSlate reported in October that some companies originally focused on Bitcoin mining are transitioning to AI infrastructure, with contracts for cloud and AI workloads, precisely because energy access has become valuable in areas like Texas. The message isn't that every miner will become an AI landlord, but that the sector's prime asset is shifting from machines to megawatts.
This pivot, however, is more difficult than it seems. AI data centers require different cooling, different network infrastructure, and different uptime guarantees. Mining can tolerate interruptions, but many AI customers will not.
The cost of rebuilding can be enormous, and competition comes from specialized data center operators with deeper relationships and financing advantages.
Yet the direction of development is clear. When energy becomes scarce, the highest-value use of a megawatt generally wins.
BlackRock's forecast isn't specifically about Bitcoin, but about the end of cheap abundance. If AI leads the US into a world where electricity demand increases rapidly while transmission remains sluggish, any business built on marginal energy economies will be crushed.
Of course, miners won't disappear. Bitcoin's incentives are designed to keep hashing power online somewhere, and the sector's mobility allows them to seek new energy sources. But the center of gravity could shift.
Regions with abundant energy production and favorable policies are expected to view miners as a stabilizing industrial burden, especially if they can offer credible curtailment. Regions welcoming hyperscalers are likely to consider miners a secondary priority.
The likely outcome is a barbell approach.
On one side are miners who integrate with grids, sign structured demand response agreements and become part of utility planning.
On the other side are miners who are converting their energy holdings into broader compute infrastructure, essentially leveraging their previous presence in energy markets to generate new income.
In any case, the easy period is coming to an end. BlackRock's warning that AI data centers could consume a huge amount of electricity in the US reminds us that the next phase of digital infrastructure will not be limited by code, but by the messy physical world of wires, permits, turbines, and heat.
What's BlackRock's main point on AI and energy?
BlackRock points out that energy demand from AI data centers is growing rapidly and electricity is an underpriced limiting factor that could have significant implications for broader energy markets.
How could changes in the energy market impact the cryptocurrency industry?
If access to electricity becomes scarce, this could put pressure on miners who rely on cheaper energy sources, jeopardizing their operating models.
What are possible adaptation strategies for Bitcoin miners?
Miners may consider shifting their focus from pure mining to hosting AI infrastructure, leveraging their existing energy infrastructure and rights.
Bitcoin (BTC)
Ethereum (ETH)
XRP (XRP)
