In 2026, several publicly listed Bitcoin miners mining companies in North America began a profound transformation, shifting from cyclical mining companies reliant on Bitcoin prices to AI data center and energy infrastructure operators, securing billions of dollars in project funding and long-term lease agreements.

Two major factors are driving this AI transformation

Following the Bitcoin halving in 2024, mining profitability continued to shrink, forcing mining companies to explore new avenues. Simultaneously, the AI industry exploded, creating a massive demand for computing power and a shortage of large-scale data centers. The industry bottleneck was not GPUs, but rather electricity, grid access, and rapid deployment infrastructure—a requirement perfectly met by the large-scale mining farms previously built by mining companies.

Mature power infrastructure, park resources, and rapid delivery capabilities are the core advantages of mining companies. While AI cloud service providers often spend years building large data centers from scratch, mining companies can significantly shorten the deployment cycle, making them the preferred partners. Currently, several leading companies have signed long-term lease contracts for hundreds of megawatts with large AI clients, with some contracts valued at tens of billions of dollars.

With this business transformation, the financing and valuation logic of mining companies has also completely changed. Previously, mining companies relied on cyclical financing methods such as equity and equipment collateral, with valuations closely tracking Bitcoin prices. Now, they generally adopt project-level debt, net leases, and take-or-pay models, leading to more stable revenue and valuations benchmarked against cash-flow steady data center real estate investment trusts (REITs). The capital market is redefining them as infrastructure assets.

The risks of this transformation are equally real and significant

On the one hand, upgrades are extremely expensive, costing millions to tens of millions of dollars per megawatt, resulting in substantial capital expenditure pressure. On the other hand, customers are highly concentrated in a few leading AI companies, and changes in demand can impact the value of long-term contracts. Furthermore, while mining companies excel in computing power and power management, they lack the operational and maintenance capabilities for large-scale AI data centers. Current high market valuations are based on successful project implementation; if execution falls short of expectations, the valuation logic may reverse.

Most AI-related projects will be launched between 2026 and 2027, with revenue currently still in the ramp-up phase. In the long run, these mining companies will diverge: some will completely transform into AI infrastructure platforms, some will retain a parallel Bitcoin mining and AI model, and companies lagging behind in the transformation will continue to be affected by the cryptocurrency industry cycle. This transformation has not only rewritten the fate of Bitcoin mining companies but has also become a typical microcosm of the revaluation of energy assets in the AI era.