Core Scientific has raised $3.3 billion to finance the construction of AI-focused data centers, marking one of the largest capital raises in the data center sector in 2026. The company, which emerged from bankruptcy in early 2024, is pivoting from Bitcoin mining to AI infrastructure by leveraging its existing power contracts, interconnection rights, and physical sites. This deal underscores investor confidence in the AI data center build-out and highlights the competitive advantage of repurposing existing infrastructure in a market constrained by interconnection delays.
The Players
Core Scientific is a former Bitcoin mining company that filed for bankruptcy in December 2022 and emerged in January 2024. The company operates multiple facilities across North America with existing power purchase agreements and grid interconnections. Core Scientific has been transitioning its business model to focus on GPU colocation and AI workloads, signing deals with hyperscalers and AI companies to lease data center capacity.
The investors and lenders in the $3.3 billion raise have not been fully disclosed in available reporting, but the size suggests a combination of institutional equity, infrastructure funds, and debt financing.
The Numbers
- Capital Raised: $3.3 billion
- Use of Proceeds: Construction of AI-focused data centers
- Existing Infrastructure: Multiple sites with power contracts and interconnection rights
- Target Market: GPU colocation for AI training and inference workloads
The financing structure likely includes a mix of senior debt, mezzanine capital, and equity to fund construction and working capital.
So What?
This deal matters because it demonstrates that repurposing existing infrastructure is a viable — and potentially faster — path to market than greenfield data center development. Core Scientific’s legacy power contracts and interconnection rights eliminate the 3–5 year queue times that are slowing new projects. In a market where speed to capacity is everything, that’s a structural advantage.
For power developers, this is both a threat and an opportunity. The threat is that companies like Core Scientific can convert existing sites faster than you can build new generation and interconnect it. The opportunity is that these conversions still need incremental power — Bitcoin mining and AI data centers have different load profiles, and retrofitting sites often requires additional capacity, upgraded cooling, and firmer power contracts.
The $3.3 billion raise also signals that investors believe the AI data center market can absorb significant new capacity. That’s a vote of confidence in sustained demand, but it also means more competition for customers and power.
What Should You Do?
If you’re a power developer, identify which Bitcoin mining sites are converting to AI and map their power needs. These sites will need incremental generation, and they will pay for speed. If you’re an investor, watch Core Scientific’s execution — the ability to convert mining infrastructure into enterprise-grade AI data centers is unproven at scale, and cooling, power density, and customer acquisition remain open questions.
Source: LetsDataScience, April 21, 2026
