In the span of a single week in April 2026, Anthropic locked in up to $65 billion in new funding commitments from two of the biggest cloud hyperscalers on the planet: $25 billion from Amazon and $40 billion from Google. These aren't venture bets from Sand Hill Road. They're infrastructure deals dressed up as investments, and they raise an uncomfortable question: what does "independence" mean when your two largest backers are direct competitors fighting over the same AI cloud market? The numbers are staggering, but the structure is what matters.

The deal, decoded

Google's commitment breaks down into two pieces: $10 billion in cash now, at a $350 billion valuation, and up to $30 billion more contingent on Anthropic hitting certain "performance milestones." Neither company has disclosed what those milestones are, who defines them, or what happens if Anthropic misses. That distinction matters. The $10 billion is a check. The $30 billion is a leash. Milestone-based funding gives Google ongoing leverage over Anthropic's strategic direction without requiring a board seat or a formal acquisition. It's soft control, the kind that doesn't show up in governance documents but absolutely shapes decision-making in the room. A week earlier, Amazon announced its own commitment: up to $25 billion on top of the $8 billion it had already invested. As part of that deal, Anthropic committed to spending more than $100 billion on Amazon Web Services over the next decade. Read that again. Anthropic is promising to return more money to Amazon than Amazon is putting in. That's not an investment in the traditional sense. That's a customer acquisition cost. Combined, Amazon and Google have now committed up to $73 billion to a single AI startup. Both are cloud providers. Both sell competing chip architectures (Trainium for Amazon, TPUs for Google). Both want Anthropic's models running on their infrastructure, not the other guy's.

Claude Code is the real story

Why are two hyperscalers tripping over each other to fund the same company? One word: Claude Code. Anthropic's agentic coding tool, which went into public beta in April 2026, has redefined what AI-assisted development looks like. This isn't autocomplete. Claude Code reads entire codebases, plans multi-step actions, executes them with real development tools, evaluates results, and iterates, all autonomously. Developers hand it a goal and step back. The growth has been explosive. Claude Code alone was reportedly approaching $1 billion in annualized revenue by late 2025, helping push Anthropic's total annual run rate to around $14 billion. That kind of traction turns a research lab into a product company overnight, and it's exactly what's driving these mega-investments. Anthropic is no longer a research lab that happens to have a product. It's a product company that happens to do research. That distinction is critical, because it changes the calculus for every investor, partner, and competitor in the space.

The OpenAI playbook

We've seen this movie before. Microsoft invested $13 billion in OpenAI starting in 2019, securing cloud exclusivity on Azure and eventually taking a 27% stake when OpenAI restructured from a nonprofit to a Public Benefit Corporation in October 2025. The trajectory is instructive. OpenAI started as a nonprofit with a mission to ensure AGI benefits all of humanity. Then came the capped-profit entity. Then the Microsoft billions. Then the restructuring. Then, in February 2026, OpenAI broke its Microsoft exclusivity by signing a $50 billion deal with AWS, prompting Microsoft to threaten legal action. At every stage, OpenAI's leadership insisted they were maintaining independence. And at every stage, the financial gravity pulled them closer to the orbits of their largest funders. Anthropic's situation is arguably more precarious. OpenAI had one dominant backer. Anthropic has two, and they're direct competitors. That might sound like it gives Anthropic more leverage, playing Amazon and Google against each other. But it also means Anthropic is making promises to both sides that may eventually prove incompatible. Anthropic has committed over $100 billion to AWS. It has simultaneously signed a deal with Google and Broadcom for "multiple gigawatts of next-generation TPU capacity" coming online in 2027. At some point, one of these partners is going to ask why their investment dollars are subsidizing the other's cloud business.

The valuation puzzle

Here's something odd: Google's $10 billion investment values Anthropic at $350 billion, the same valuation the company had back in February during its $30 billion Series G round led by GIC and Coatue (which actually valued Anthropic at $380 billion post-money). In the two months between those rounds, Claude Code exploded. Anthropic's revenue trajectory steepened. Secondary market valuations surged past $1 trillion, with desperate buyers on platforms like Forge Global and Hiive bidding up shares to prices that would have seemed absurd weeks earlier. Some traders reported Anthropic overtaking OpenAI in secondary market pricing. So why is Google getting in at a flat $350 billion? The most likely answer: leverage. When you're writing a $10 billion check with $30 billion more on the table, you set the price. The hyperscalers aren't passive investors competing in an auction. They're strategic partners with something Anthropic desperately needs, compute, and they're pricing accordingly.

What "independence" actually means

Anthropic was founded in 2021 by former OpenAI researchers, including Dario and Daniela Amodei, with an explicit pitch: we'll be the safety-focused alternative, the "third option" in a market dominated by OpenAI and Google DeepMind. The company positioned itself as the responsible counterweight to the race dynamics it saw destroying OpenAI's original mission. That narrative is getting harder to maintain. When your two largest backers account for $73 billion in committed funding and your business depends on their chips and cloud infrastructure, "independence" becomes a matter of degree, not kind. This doesn't mean Anthropic has been captured. The company still controls its own board. It still sets its own research agenda. It recently held back the public release of its Mythos model over safety concerns, a decision that a purely profit-driven entity might not have made. These are real signals of institutional independence. But institutional independence and financial independence are different things. Anthropic is now locked into decade-long infrastructure commitments to both Amazon and Google. Its future compute roadmap depends on chips and capacity that these two companies control. And with an IPO potentially on the horizon, likely as soon as October 2026 with a target north of $60 billion, the pressure to deliver returns will only intensify.

Why builders should care

If you're building on Claude, does any of this matter? Arguably, yes. When your model provider is financially entangled with multiple cloud platforms, your data flows through those platforms too. Enterprise customers choosing between AWS Bedrock and Google Cloud for their Claude deployments aren't just making a technical decision. They're choosing which hyperscaler's ecosystem gets deeper access to their usage patterns, their workloads, and their dependency graphs. More practically, the milestone-based funding structure means Anthropic's product roadmap may be shaped by what Google and Amazon define as "performance targets." If those targets emphasize cloud consumption metrics over, say, safety research milestones, the incentives tilt in a particular direction. None of this is unique to Anthropic. The entire frontier AI industry is converging on the same model: research labs need compute, hyperscalers have compute, and the resulting financial entanglements blur the line between partner and subsidiary. But Anthropic is the most visible test case for whether a company can take billions from competing interests and still chart its own course.

The trillion-dollar question

Anthropic's secondary market valuation recently crossed $1 trillion. An IPO could make it one of the largest public offerings in history. Claude Code is rewriting how software gets built. By any conventional measure, the company is winning. But winning and staying independent aren't the same thing. Google didn't invest $40 billion out of altruism. Amazon didn't commit $25 billion to advance the mission of AI safety. These are strategic bets by companies that want Anthropic's technology locked into their ecosystems, and they're structured in ways that create long-term dependency without requiring formal control. The question isn't whether Anthropic can maintain its independence today. It's whether the financial architecture being built around it will allow independence to survive the next five years. If history is any guide, the answer isn't encouraging.

References