Full Breakdown
Google Plans Up to $40 Billion Investment in Anthropic Amid AI Infrastructure Race
4/28/2026, 5:03:32 AM
Deal Overview
Alphabet will invest $10 billion in Anthropic at a $350 billion valuation, with a further $30 billion tied to performance milestones. The pact grants Anthropic up to five gigawatts of Google Cloud compute, plus additional gigawatt-scale capacity from Broadcom and CoreWeave.
Background & Context
Anthropic, founded in 2021 by former OpenAI researchers, has seen its annual revenue run-rate rise from about $9 billion at the end of 2025 to over $30 billion in early 2026, driven by Claude Code and Claude Cowork. Google bought a $300 million stake in February 2023, added $2 billion in October 2023, and now holds a 14 % share.
Data & Statistics
Investment: $10 bn upfront, up to $30 bn contingent; Valuation: $350 bn, with other reports of $380 bn or $800 bn; Compute: 5 GW from Google Cloud (starting 2027), ~1 GW via Amazon chips, plus Broadcom/CoreWeave capacity; Revenue: >$30 bn run-rate, up from $9 bn (2025); Planned U.S. data-center spend: $50 bn.
Why It Matters
The pact highlights the capital-intensive nature of generative AI and the strategic value of compute. By linking most funding to performance, Alphabet secures long-term TPU and cloud usage while sharing risk. The parallel Amazon commitment illustrates a broader “circular-deal” trend that could concentrate AI infrastructure power.
Official Statements & Responses
Anthropic’s chief executive said users consider Claude essential and that the startup must expand its infrastructure to meet rising demand. Neither Google nor Amazon provided additional remarks.
Criticism & Opposition
U.S. defense officials have labeled Anthropic a potential supply-chain risk, prompting a legal challenge. Analysts warn that “circular deals” may raise antitrust concerns by tying financing to infrastructure provision, limiting competition for smaller AI firms.
Conflicting Reports & Gaps
Sources disagree on Anthropic’s valuation—some cite $350 bn, others $380 bn, and a few suggest a potential $800 bn figure. The specific performance targets that would unlock the $30 bn tranche are not disclosed, and rollout timelines for the full gigawatt compute capacity remain vague.
Verbatim Quotes
- “Our users tell us Claude is increasingly essential to how they work, and we need to build the infrastructure to keep pace with rapidly growing demand,” — Dario Amodei, CEO, Anthropic
- “Google is ready to make another multi-billion-dollar investment in Anthropic, as the Claude developer continues to seek computing power to sustain its growth.” — Bloomberg report
- “Alphabet is reportedly planning to invest up to $40 billion in Anthropic, significantly expanding its commitment to one of the leading developers of large language models, according to a CNBC report.” — CNBC report
- “Investment deepens Big Tech race for AI infrastructure and signals rising value of compute at scale The AI race is no longer just about models.” — WayA Media analysis
What’s Next
Anthropic’s next milestones include scaling its gigawatt-level compute fleet and converting the $30 bn tranche into cash as model adoption and revenue grow. An IPO is rumored for late 2026, testing market appetite for AI valuations. Google’s cloud and TPU roadmap will evolve to support the partnership, while regulatory scrutiny of “circular deals” may shape future investment structures.
