Google to invest up to $40B in Anthropic in cash and compute

Google is reportedly preparing a major investment of up to $40 billion in Anthropic to strengthen the startup’s expanding computing infrastructure, according to Bloomberg. The initial phase includes a $10 billion commitment based on a $350 billion valuation, with an additional $30 billion expected to be released if Anthropic achieves specific performance milestones.

This investment announcement follows the recent rollout of Anthropic’s newest AI model, Mythos, which has been introduced to a select group of partners. The company describes Mythos as its most advanced system so far, particularly highlighting its potential in cybersecurity. However, due to concerns around possible misuse, access has been tightly controlled while testing continues with trusted organizations. Despite these precautions, the model has reportedly already been accessed without authorization, and its operational costs at scale are expected to be significant.

As competition in artificial intelligence intensifies, access to large-scale computing resources has become a critical factor. OpenAI, for instance, has been aggressively securing infrastructure through massive deals involving cloud platforms, chip manufacturers, and energy providers, including a recent expansion with Cerebras.

Anthropic, meanwhile, has also been racing to expand its infrastructure capabilities. The company recently faced criticism over usage limits for its Claude platform, prompting it to pursue multiple infrastructure agreements. Earlier this month, it partnered with CoreWeave for additional data center capacity. It also secured a further $5 billion investment from Amazon as part of a broader deal that could see Anthropic invest up to $100 billion over time to secure approximately 5 gigawatts of compute power.

Although Google competes directly in the AI model space, it remains a crucial infrastructure partner for Anthropic. The company depends heavily on Google Cloud services, particularly its tensor processing units (TPUs), which are specialized chips designed for AI workloads and are considered strong alternatives to those produced by Nvidia.

The collaboration between Google and Anthropic is not new. Earlier developments included a partnership involving Broadcom, which works with Google to design custom AI chips. This agreement is expected to provide Anthropic with access to several gigawatts of TPU-based computing capacity starting in 2027, with filings indicating an estimated 3.5 gigawatts.

With the latest investment, Google Cloud is set to further expand its support by delivering an additional 5 gigawatts of computing capacity over the next five years, with the potential for further scaling.

Anthropic’s valuation was estimated at $350 billion as of February, but investor interest has reportedly pushed expectations toward $800 billion or more. The company is also said to be exploring the possibility of going public as early as October.

Voice Of Osiz

At Osiz, we see Google’s proposed $40B investment in Anthropic as a defining moment in the evolution of AI infrastructure. This move highlights a critical industry shift—where access to high-performance compute is becoming just as valuable as the AI models themselves. The collaboration underscores how strategic partnerships between AI innovators and cloud providers are shaping the next wave of scalable, secure, and high-impact AI solutions. With Anthropic focusing on advanced models like Mythos and Google enabling massive compute capabilities, the foundation for enterprise-grade AI adoption is being strengthened.

However, this development also brings attention to key challenges, including cost optimization, responsible AI deployment, and infrastructure scalability. Businesses must now think beyond model development and prioritize robust ecosystems that support performance, security, and compliance. At Osiz, we help enterprises navigate this rapidly evolving landscape by delivering end-to-end AI development solutions—from model design to scalable deployment powered by cutting-edge infrastructure. The future of AI belongs to those who can seamlessly integrate innovation with infrastructure—and this investment is a clear signal of that direction.

Source: Techcrunch.com