According to Inc., CoreWeave’s $9 billion acquisition bid for crypto miner Core Scientific was terminated yesterday after shareholders voted against the deal despite board support. The proposed acquisition, first announced in July, would have given CoreWeave ownership of Core Scientific’s data centers and energy capacity to meet rising AI infrastructure demand. Major shareholder Two Seas Capital and advisory firm Institutional Shareholder Services both opposed the deal, citing valuation concerns and arguing it “materially undervalues” Core Scientific. This marks the second failed acquisition attempt between the companies this year, following Core Scientific’s rejection of an all-cash offer from CoreWeave in June 2024. The rejection comes as analysts suggest shareholders believe their companies deserve higher valuations based on current AI company stock multiples.
Table of Contents
The AI Infrastructure Gold Rush Intensifies
The failed acquisition reveals just how desperate AI infrastructure companies have become for data center capacity. CoreWeave, which specializes in cloud computing for artificial intelligence workloads, already rents capacity from Core Scientific, but ownership would have provided strategic control over critical infrastructure. The AI boom has created unprecedented demand for high-performance computing resources, particularly GPU clusters needed for training and running large language models. What makes this situation particularly telling is that Core Scientific represents exactly the type of infrastructure play that AI companies are scrambling to secure – existing data centers with substantial power contracts and cooling capacity that can be repurposed from cryptocurrency mining to AI computation.
Shareholder Activism in the AI Era
The rejection highlights a significant shift in shareholder behavior during technology transitions. Historically, acquisition targets in emerging sectors often accepted buyout offers from larger players. However, the current environment shows that proxy voting and shareholder activism have become powerful tools for capturing what investors perceive as full value. As CoreWeave stated in their official response, they must now navigate a landscape where target companies’ shareholders are increasingly sophisticated about AI infrastructure valuations. This creates a challenging environment for strategic acquisitions, potentially forcing companies to pursue more expensive greenfield construction or complex partnership structures instead of straightforward acquisitions.
AI Valuation Bubble Concerns Surface
The analyst commentary about “AI trade froth” versus “actual economic value” points to growing concerns about sustainable valuations in the AI infrastructure space. While CoreWeave and similar companies have seen remarkable growth, shareholders of acquisition targets are now demanding premiums that reflect not just current operations but anticipated future AI-driven revenue. This creates a fundamental tension – acquiring companies want to buy assets before prices rise further, while target company shareholders want to capture the full upside of the AI transformation. The situation mirrors early internet boom dynamics where infrastructure companies became overvalued based on projected demand that took years to materialize. As analysis suggests, this failed deal may actually benefit Core Scientific by allowing them to capture more value independently.
Strategic Implications for AI Infrastructure
Looking forward, this failed acquisition signals several important trends. First, the scramble for AI-ready data center capacity will likely intensify, with companies exploring alternative strategies including build-to-suit arrangements, joint ventures, and longer-term lease agreements. Second, we may see more creative deal structures that allow selling shareholders to participate in future upside through earn-outs or stock consideration rather than fixed valuations. Third, the rejection suggests that infrastructure owners are becoming increasingly sophisticated about their strategic value in the AI ecosystem. Companies like Core Scientific now understand they’re not just selling data centers – they’re selling access to the computational foundation of the next technological revolution, and they’re pricing accordingly.
