Anthropic, the AI company behind the Claude family of models, is on track to receive up to $40 billion from Google in what would rank as one of the largest single technology investments ever made. Bloomberg reported in late April 2026 that the deal includes $10 billion in immediate cash at a reported $350 billion valuation, with up to $30 billion more contingent on performance milestones that have not been publicly disclosed.
The commitment would place Anthropic at the center of a spending war between the world’s largest cloud providers. Amazon has separately pledged $5 billion now and up to $20 billion more in future investment, alongside a deal committing Anthropic to more than $100 billion in Amazon Web Services spending over the next decade, according to the Associated Press. No other AI startup is drawing tens of billions simultaneously from two direct competitors.
The hardware backbone
The partnership extends well beyond an equity check. A Broadcom Form 8-K filed with the SEC describes agreements with Google for custom tensor processing units (TPUs) and networking hardware stretching through 2031. The filing confirms that Broadcom, Google, and Anthropic have expanded a three-way collaboration, with Broadcom supplying advanced AI accelerators to support Anthropic’s scaling plans.
The earliest event in that filing is dated April 6, 2026, weeks before Bloomberg published the broader financial terms. That sequence suggests the infrastructure agreements were hammered out first, forming the foundation on which the investment was built. For Google, tying custom silicon to a specific AI partner is a bet that its hardware roadmap and Anthropic’s model development will evolve together for at least five years.
A $350 billion price tag in context
If the reported valuation holds, Anthropic would rank among the most valuable private companies on the planet. For comparison, OpenAI closed a funding round in late 2024 that valued it at roughly $157 billion, according to Reuters. Anthropic’s figure, less than two years later, reflects how rapidly investor appetite for frontier AI companies has escalated.
Private valuations deserve scrutiny, though. They are set by negotiated deal terms, not by open-market trading, and can shift dramatically between rounds. Neither Google nor Anthropic has confirmed whether the $350 billion figure is pre-money or post-money, what liquidation preferences or governance rights attach to Google’s shares, or how much downside protection Google secured. Without those details, the number signals market enthusiasm more than it does a precise measure of Anthropic’s worth.
Anthropic has not publicly disclosed annual revenue figures, making it difficult to judge the valuation against conventional financial metrics. The company has reportedly been generating revenue through API access and enterprise contracts on both Google Cloud and AWS, but the scale of that business relative to a $350 billion price tag remains an open question.
What has not been confirmed
Several critical elements remain unverified by the companies involved. As of early May 2026, neither Google nor Anthropic has issued a public statement or regulatory filing detailing the $40 billion investment structure, the benchmarks tied to the $30 billion contingent tranche, or the deployment timeline. Bloomberg’s figures are sourced to people familiar with the matter, and final terms sometimes shift between initial reports and signed agreements.
The Broadcom filing, while legally binding as a regulatory disclosure, does not specify exact hardware allocations, delivery schedules, or financial terms for the Anthropic-related work. One passage in the publicly available document appears to reference Anthropic’s expanded role beginning “in 2” followed by text that is cut off, likely an artifact of redaction or a formatting error in the SEC submission. The incomplete phrase leaves the precise start date of the expanded collaboration ambiguous.
On the Amazon side, the $100 billion AWS commitment and the 5-gigawatt Trainium computing capacity figure come from AP reporting rather than from any SEC filing or official Amazon disclosure. The relationship between the two competing investments raises unanswered questions: whether Anthropic must prioritize one cloud for certain workloads, whether governance rights could give either investor influence over Anthropic’s product roadmap, and how two rival platforms intend to share a single, strategically vital AI partner.
The regulatory question no one has answered
Deals of this magnitude typically attract attention from antitrust regulators, and this one involves unusual dynamics. Two of the three dominant U.S. cloud providers are pouring capital into the same startup, potentially shaping which AI models enterprise customers can access and on what terms. The Federal Trade Commission has already been scrutinizing large AI partnerships since launching a formal inquiry into generative AI investments in January 2024. Whether the Google and Amazon commitments to Anthropic trigger deeper review, or whether their competing nature actually reduces concentration concerns, is a question regulators have not yet addressed publicly.
Why the dual-cloud strategy matters
Most AI startups align tightly with a single cloud provider. Anthropic’s ability to extract massive capital commitments from both Google Cloud and AWS breaks that pattern and gives the company unusual leverage in negotiations over pricing, infrastructure access, and independence.
For Google, the investment is defensive as much as it is offensive. Microsoft has backed OpenAI with commitments widely reported to exceed $13 billion, and Amazon’s Anthropic deal already gave AWS a powerful model to offer enterprise customers. By investing up to $40 billion and securing a custom chip pipeline through Broadcom, Google is ensuring that its cloud platform has access to one of the two leading families of frontier AI models rather than ceding that ground entirely to rivals.
For the broader AI industry, the capital flowing into Anthropic underscores a stark reality: building and running frontier models now requires infrastructure spending measured in tens of billions of dollars. That cost structure favors companies with deep-pocketed backers and raises hard questions about whether smaller AI labs can compete at the same scale, or whether the industry is consolidating around a handful of model providers backed by the largest cloud platforms.
The milestones that will define the deal
Market watchers will be looking for follow-up SEC filings, earnings call commentary from Google and Amazon, or a joint announcement that clarifies how much capital ultimately changes hands and what constraints Anthropic accepts in return. The performance milestones tied to the $30 billion contingent tranche are particularly important: if they are linked to model capability benchmarks, revenue targets, or compute scaling goals, they will reveal what Google believes Anthropic must achieve to justify the full investment.
Until those disclosures arrive, the strongest evidence points to a partnership that is real, long-dated, and anchored in hardware commitments extending through 2031. The precise valuation and funding mechanics remain provisional. But the reported terms alone make one thing clear: the race for computational scale in AI has entered a phase where the price of a seat at the table is measured not in billions, but in tens of billions. The biggest technology companies in the world are willing to pay it.
