Anthropic Secures AI Compute Power with Google, Broadcom

Anthropic has forged a significant strategic alliance with tech giants Google and Broadcom, securing a substantial supply of advanced artificial intelligence processing units. This multi-year agreement is designed to guarantee access to next-generation tensor processing units (TPUs) with multi-gigawatt capacity, slated for deployment in 2027. The move is critical for Anthropic as it aims to meet the escalating global demand for its sophisticated AI models, including the widely adopted Claude series, by bolstering its infrastructure for both training and inference.

The partnership addresses a core bottleneck in the AI industry: the availability of specialized hardware. With the AI sector experiencing exponential growth, projected to reach hundreds of billions in market value within the next decade, securing compute power is paramount. This deal positions Anthropic to scale its operations effectively, supporting its rapidly expanding customer base, which now includes over 1,000 businesses each spending more than $1 million annually on its services. This surge in enterprise adoption highlights the increasing reliance on advanced AI for business operations.

Beyond hardware, Anthropic also revealed a dramatic acceleration in its financial performance. The company's run-rate revenue has now surpassed the $30 billion mark, a substantial leap from $9 billion recorded at the close of 2025. This impressive growth underscores the market's strong appetite for Anthropic's AI solutions and its successful commercialization strategy. The company's CFO, Krishna Rao, emphasized that this expansion is a direct result of sustained customer trust and the company's commitment to building robust infrastructure.

The collaboration with Google and Broadcom is multifaceted. Broadcom, in a separate regulatory filing, confirmed a long-term commitment to supply Google with future generations of TPUs, specifically engineered for data center AI workloads. This intricate supply chain arrangement ensures Anthropic gains access to approximately 3.5 gigawatts of TPU-driven computational power, a figure that underscores the scale of this strategic commitment and its implications for large-scale AI deployment.

This development also aligns with broader national initiatives to bolster domestic computing infrastructure. A significant portion of the new computing resources will be situated within the United States, reinforcing Anthropic's pledge to invest heavily in American digital infrastructure. This strategic placement is crucial for data sovereignty, security, and reducing latency for its North American clientele, further solidifying its position as a key player in the U.S. AI ecosystem.

While Anthropic focuses on scaling its AI capabilities, the company is also navigating a complex regulatory and public perception environment. Recent reports have touched upon its past disagreements with governmental bodies regarding AI usage policies and its classification as a supply-chain risk. Concurrently, the company is reportedly preparing for a potential initial public offering (IPO) as early as October, engaging in preliminary discussions with financial institutions, positioning itself as a formidable contender in the competitive AI market alongside rivals like OpenAI.