Google to invest up to $40bn in Anthropic at $350bn valuation

TL;DR:

  • Alphabet will invest up to $40 billion in Anthropic — $10 billion now in cash at a $350 billion valuation, plus $30 billion if performance targets are met.
  • The deal lands days after Amazon committed up to $25 billion to the Claude maker, leaving the same labs hyperscaler-funded by both Google and Amazon simultaneously.
  • For UK enterprise buyers, the practical implication is that Anthropic’s compute pipeline is now backed across two of the three big clouds — reducing single-vendor concentration risk in their procurement model.

Google parent Alphabet will invest up to $40 billion in Anthropic, with $10 billion committed now in cash at a $350 billion valuation and a further $30 billion contingent on performance milestones, according to Reuters. The capital is earmarked for a major expansion of Anthropic’s computing capacity, the company said on Friday.

The commitment arrives just days after Amazon announced it would invest up to $25 billion in the same company. The combined effect is that Anthropic is simultaneously being backed at hyperscale by two competing cloud providers — a structurally different position from the OpenAI–Microsoft pairing, where compute and capital sit predominantly with one cloud.

Run-rate, valuations and a bigger bidding war

Anthropic’s annual run-rate revenue passed $30 billion this month, up from roughly $9 billion at the end of 2025, with Reuters attributing much of that growth to the take-up of Claude Code among developers. The company raised $30 billion in February at a $380 billion post-money valuation, and recent reports suggest secondary offers from venture firms valuing it as high as $800 billion.

To absorb the demand, Anthropic has lined up multiple compute deals: multi-year agreements with chipmaker Broadcom and cloud-infrastructure firm CoreWeave earlier this month, and roughly 1 GW of capacity via Amazon’s chips by year-end. Last year the company said it would deploy $50 billion building US data-centre capacity to deploy and train its own models.

Looking forward

For UK businesses already running Claude through AWS Bedrock or considering Anthropic via Vertex AI, the dual hyperscaler funding clarifies a procurement question that has been ambiguous all year: neither AWS nor Google Cloud has reason to throttle Anthropic access in favour of an alternative model, because both are now equity holders. That reduces the risk of a sudden contractual pivot.

The same setup raises a subtler question for the broader market. If frontier labs increasingly raise capital from multiple hyperscalers rather than aligning with one, the assumption that “AI vendor lock-in” maps cleanly to “cloud lock-in” gets weaker. The UK firms still building procurement frameworks on that assumption may want to revisit them.

Share this article