Skip to main content
GPUBeat Inference & Serving CoreWeave CEO Michael Intrator Sells $30.2M…

CoreWeave CEO Michael Intrator Sells $30.2M in Stock Amid AI Expansion

Michael Intrator's recent stock sale raises questions about CoreWeave's strategy as the company expands its AI cloud capabilities amid growing competition from tech giants.

CoreWeave stock sale and AI cloud expansion — Michael N. Intrator, CoreWeave, Inc.
CoreWeave CEO Michael Intrator Sells $30.2M in Stock Amid AI Expansion Source: GPUBeat

Michael N. Intrator, CEO of CoreWeave, Inc., has divested approximately $30.2 million worth of Class A Common Stock, selling 307,692 shares on May 19, 2026. This sale took place while CoreWeave shares traded at $107.58, marking a significant 50% increase in value over the past six months, despite some recent fluctuations.

Intrator sold 199,999 shares directly, with prices ranging from $94.95 to $101.32 each. Omnadora Capital LLC, through which he holds securities, sold 107,693 shares within the same price bracket. These transactions were carried out under a Rule 10b5-1 trading plan that Intrator established in November 2025. After these sales, Intrator retains 4,266,501 shares of Class A Common Stock, while his indirect holdings through Omnadora have been reduced to zero.

This stock sale comes as CoreWeave gears up for significant expansion, underscored by the successful closing of a $3.1 billion delayed draw term loan facility (DDTL 5.0). This financing is particularly noteworthy as the first publicly syndicated loan backed by high-performance computing infrastructure, earning solid ratings of Ba2 from Moody’s and BB+ from Fitch. The funds will aid CoreWeave's efforts to grow its AI cloud platform, an increasingly competitive area.

CoreWeave's Growth Amidst Competition

In addition to securing financing, CoreWeave is enhancing its offerings with the launch of CoreWeave Sandboxes, which provide secure environments for training AI models. Users can access these sandboxes through the CoreWeave Kubernetes Service or serverless configurations via Weights & Biases. CoreWeave also achieved a significant milestone by recording the top inference speed for the Kimi K2.6 model, processing 205 tokens per second at a cost of $0.7 per million tokens, according to data from Artificial Analysis.

See also  Navigating GPU Costs: The Hidden Expenses of Multi-Cloud Inference

However, the space is evolving. CoreWeave may face challenges from a new AI cloud initiative backed by Google and Blackstone, which plans to bring substantial capacity online by 2027. This venture is supported by a $5 billion investment and aims to utilize Google’s TPU technology to provide alternative access to Tensor Processing Units. Analysts are divided on CoreWeave's outlook; Evercore ISI maintains an Outperform rating, citing strong demand for AI, while Bernstein SocGen Group expresses concerns, issuing an Underperform rating due to fears over rising competition from hyperscalers.

Financial Outlook and Analyst Sentiments

Despite impressive recent revenue growth of 130% year-on-year, CoreWeave's financials reveal ongoing challenges, with an earnings per share (EPS) of -$3.15, indicating unprofitability. The company's current market capitalization is $54.45 billion, and analyses suggest it may be undervalued at current price levels. InvestingPro has identified 16 additional ProTips for investors, offering a wealth of actionable insights.

As CoreWeave navigates this complex environment, Intrator's stock sale raises questions about the company’s internal strategies and future directions. The looming competition from major players could further influence the AI cloud market, making it essential for CoreWeave to solidify its position and builds innovation to stay ahead.

The changing dynamics emphasize the importance for companies like CoreWeave to adapt and respond effectively to funding opportunities and competitive pressures in the rapidly growing AI sector.

GD

GPUBeat Desk

Desk · joined 2026

GPUBeat Desk covers AI infrastructure — chips, foundation models, inference economics, datacenter buildouts, and the geopolitics of compute.