Google and Blackstone partner to launch new AI cloud company

San Jose, California (US), May 19, 2026, BNN Web Staff: Tech giant Google and investment firm Blackstone have teamed up to establish a new artificial intelligence cloud infrastructure company aimed at competing in the rapidly expanding AI computing market.

The new venture is expected to challenge firms such as CoreWeave by using Google’s specialised Tensor Processing Units (TPUs), instead of relying primarily on Nvidia-powered infrastructure that currently dominates the sector.

According to reports, Blackstone plans to invest nearly $5 billion in equity capital into the venture and is expected to hold a majority stake in the company.

The partnership marks Google’s most ambitious effort yet to commercially expand the use of its in-house AI chips and strengthen its position in the cloud infrastructure business, where demand for large-scale computing power has surged due to the rapid growth of generative AI technologies.

The proposed AI cloud company aims to bring around 500 megawatts of computing capacity online by 2027, reflecting the massive scale of infrastructure now required to support advanced AI models and enterprise-level applications.

Blackstone has significantly increased its investments in artificial intelligence and data centre infrastructure in recent years. The company already holds investments in AI-related firms including OpenAI and Anthropic, alongside its stake in CoreWeave.

The move comes as corporations worldwide rapidly adopt AI-driven tools, increasing the need for high-performance cloud systems capable of training and running large AI models.

Earlier this year, Sundar Pichai highlighted Google’s strong AI-driven growth, stating that demand for AI products and infrastructure had accelerated significantly across its cloud business. He noted that Google Cloud revenue had crossed $20 billion for the first time, driven largely by enterprise AI adoption and growing use of Gemini AI products.

Pichai also revealed that Google’s AI systems are now processing more than 16 billion tokens per minute through direct customer API usage, reflecting the rapidly increasing scale of AI deployment globally.