BlackRock Sees AI Power Demand Disrupting Bitcoin Mining Model

TLDR

Table of Contents

BlackRock projects AI data centers may use up to 24% of US power by 2030.
AI needs constant electricity, unlike Bitcoin mining’s flexible usage model.
Grid bottlenecks may favor AI over mining due to political and economic value.
Bitcoin miners may shift to hosting AI infrastructure or sign curtailment deals.

BlackRock has issued a warning that artificial intelligence (AI) and Bitcoin mining are heading toward a power conflict. In its 2026 Global Outlook, the firm predicts AI data centers could consume up to 24% of US electricity by 2030. This growing energy demand could push Bitcoin miners out of key markets as grid access becomes scarce and politically sensitive.

AI’s Soaring Energy Demand Reshapes the Power Market

In its 2026 Global Outlook, BlackRock said AI is no longer just a software story but an energy-intensive sector. The investment firm projects AI data centers could consume up to 24% of US electricity by 2030. This figure far exceeds current estimates by agencies such as the US Department of Energy and the Lawrence Berkeley National Laboratory.

NEW: BLACKROCK WARNS AI’S ENERGY DEMAND COULD STRAIN U.S. ELECTRICITY SUPPLY pic.twitter.com/D6BT8oXp3W
— The Wolf Of All Streets (@scottmelker) January 11, 2026

The expected surge in electricity use comes from the capital-intensive nature of AI infrastructure. AI systems require high uptime, massive compute power, and non-stop energy. As a result, data centers supporting AI workloads are locking in long-term contracts for guaranteed power delivery. This shift puts pressure on grids already adjusting to rapid demand from electric vehicles and other industrial uses.

Bitcoin Miners Lose Their Energy Edge

Bitcoin miners have historically built operations in areas with low-cost, interruptible power. They have been able to pause operations during grid stress and benefit from demand-response programs, such as those run by Texas’s ERCOT. In some cases, miners have even earned millions by powering down during peak demand periods.

However, this flexibility is no longer enough as utilities prioritize loads that support economic and national goals. AI data centers, with their consistent and high energy requirements, are seen as more essential. A miner’s ability to curtail usage may no longer be attractive compared to an AI data center’s promise of long-term jobs and strategic value.

BlackRock warns that as electricity becomes a scarce resource, lawmakers and regulators may favor AI infrastructure. The perception that AI contributes to productivity and national competitiveness could reduce political support for crypto mining.

Grid Constraints Shift the Business Model

According to BlackRock, miners once benefited from fast deployment and access to underused generation. But as power access becomes a regulatory hurdle, fast deployment loses its advantage. Grid interconnection queues, infrastructure delays, and permitting processes are now the main challenges.

Regions such as Texas and Northern Virginia are already seeing power demand outpace grid capacity. New industrial projects may struggle to access substations or transmission lines, regardless of how quickly they can be built. In this environment, miners are less likely to be granted priority access over AI operators.

The firm suggests that the value in mining is shifting from computational hardware to control over megawatts. Land with existing substations and power rights could now serve a more profitable use by hosting AI infrastructure.

Miners Pivot to Hosting AI or Emphasize Flexibility

Some Bitcoin mining firms are already exploring alternative business models. By transforming their sites into AI compute facilities, they can offer land, power, and infrastructure to cloud providers. CryptoSlate reported that several miners have signed contracts to host AI workloads, signaling a shift toward stable, contracted revenue.

Others may continue mining while integrating further into grid planning. By signing structured demand-response agreements, they can still present themselves as valuable, flexible industrial loads. Duke University research suggests the existing grid can handle more load if it can be curtailed when needed.

BlackRock suggests both models may co-exist. Miners that adapt to tighter grid controls or pivot to AI hosting could stay competitive. But those that rely on cheap and flexible power without adjustment may struggle in a changing energy landscape.