Google commits $40 billion to Anthropic in largest AI investment on record
Tiered deal structure—$10B now, $30B conditional—pairs equity with guaranteed 5-gigawatt compute allocation, dwarfing Microsoft-OpenAI and signaling strategic pivot from horizontal LLM competition to infrastructure control.
Google will invest up to $40 billion in Anthropic across a tiered structure that combines immediate equity, performance-linked tranches, and guaranteed cloud compute—the largest single financial commitment to an AI startup outside Microsoft’s OpenAI partnership. The deal, announced 24 April, delivers $10 billion in cash now at a $350 billion valuation, with another $30 billion contingent on performance milestones, according to Bloomberg.
The commitment arrives days after Amazon pledged $5 billion immediately and up to $20 billion more tied to commercial milestones, expanding its prior $8 billion stake, per Amazon. Combined, the dual deals deliver roughly $65 billion in pledged capital and 10 gigawatts of reserved training capacity to Anthropic within a single week—infrastructure firepower that reframes the economics of frontier model development ahead of an expected 2026 IPO.
Infrastructure as strategic moat
Google Cloud will provide 5 gigawatts of capacity over the next five years, with room to scale further, TechCrunch reports. The allocation centers on Google’s Tensor Processing Units (TPUs), which Anthropic has used for several years. “Anthropic’s choice to significantly expand its usage of TPUs reflects the strong price-performance and efficiency its teams have seen,” said Thomas Kurian, CEO of Google Cloud, in a statement cited by PYMNTS.
“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, Anthropic CEO
The compute commitment effectively locks Anthropic into Google’s hardware ecosystem for the next half-decade while offering predictable capacity in a market where GPU scarcity has become the binding constraint on model development. OpenAI expects compute access to hit 30 gigawatts by 2030, according to Axios—a target that underscores the scale of infrastructure required to train next-generation models.
Revenue traction drives valuation
Anthropic’s annualized revenue run rate hit $30 billion in April, up from $1 billion in January 2025, per The Next Web. The acceleration reflects enterprise adoption of Claude, Anthropic’s flagship model, which has gained traction in sectors prioritizing interpretability and safety—attributes central to the company’s constitutional AI framework. The revenue surge provides commercial justification for the $350 billion valuation, though the run-rate figure represents recent momentum rather than normalized full-year results.
| Deal | Cumulative Capital | Compute Allocation |
|---|---|---|
| Google-Anthropic | $40B (conditional) | 5 GW over 5 years |
| Amazon-Anthropic | $33B (conditional) | $100B AWS commitment |
| Microsoft-OpenAI | ~$13B | Undisclosed |
Google’s existing stake in Anthropic stands at approximately 14%, acquired through more than $3 billion in prior investments. The stake is contractually capped at 15% and carries no voting rights, no board seats, and no board observer privileges—a structure designed to preempt Antitrust scrutiny while preserving commercial flexibility.
Strategic admission: Gemini trails in enterprise
The scale of Google’s commitment reflects competitive pressure in the segment that generates recurring revenue: enterprise deployments. Claude is outperforming Gemini in commercial accounts, forcing Google to secure access to Anthropic’s technology rather than rely solely on internal models, according to deal analysis. The partnership allows Google Cloud to bundle Claude alongside Gemini, offering customers model choice while monetizing inference workloads regardless of which system they select.
The arrangement mirrors Microsoft’s OpenAI strategy: act as infrastructure provider and distribution channel while hedging against the risk that internal models fail to achieve commercial parity. For Anthropic, the deal trades equity dilution for guaranteed compute and access to Google’s enterprise sales apparatus—a trade-off that accelerates go-to-market velocity but embeds dependency on a competitor.
Anthropic committed to spend more than $100 billion over the next ten years on AWS technologies as part of its expanded Amazon partnership, creating parallel dependencies on two competing hyperscalers. The dual-track strategy—sometimes termed “polyamorous” hyperscaler relationships—allows model developers to negotiate better terms but raises questions about long-term control over training infrastructure and commercial distribution.
Antitrust implications expand
Google now operates simultaneously as investor, infrastructure gatekeeper, and competitor to Anthropic—a configuration that regulators have flagged in prior reviews of hyperscaler-AI lab partnerships. The FTC and DOJ are likely to examine whether the compute allocation functions as a vertical foreclosure mechanism, limiting Anthropic’s ability to shift workloads to competing clouds or negotiate independently with hardware suppliers, according to TechStory.
The European Commission has opened preliminary inquiries into Microsoft-OpenAI and Amazon-Anthropic deals under merger control frameworks, focusing on whether minority investments that include compute exclusivity should trigger formal review thresholds. Google’s deal—larger in absolute terms and paired with multi-year infrastructure lock-in—fits the same pattern.
- $40B commitment dwarfs all prior AI investments outside Microsoft-OpenAI, with $10B delivered now and $30B conditional on milestones
- 5-gigawatt compute allocation over five years locks Anthropic into Google’s TPU ecosystem
- Anthropic’s $30B annualized revenue run rate validates enterprise demand for constitutional AI approach
- Deal structure embeds Google as both infrastructure provider and commercial competitor, raising antitrust questions
What to watch
Monitor FTC and European Commission filings for formal inquiries into the compute allocation terms, particularly whether exclusivity clauses limit Anthropic’s ability to train on non-Google hardware. Track Anthropic’s AWS spending against the $100 billion commitment to Amazon—divergence would signal strategic prioritization of one hyperscaler over the other. Watch for enterprise customer disclosures from Google Cloud and AWS that break out Claude-related revenue, which would clarify whether Anthropic is driving new workloads or cannibalizing Gemini deployments. Finally, observe whether Microsoft adjusts OpenAI terms in response—the $13 billion cumulative investment now looks modest by comparison, potentially forcing renegotiation to maintain competitive parity in model access and compute allocation.