Nvidia’s $70 Bn AI Bet May Face a $2 Tn Reality Check as Anthropic Eyes IPO

Nvidia’s historic artificial intelligence windfall is transforming the semiconductor giant into one of Silicon Valley’s most powerful AI capital engines. The chipmaker has committed about $70 Bn to direct equity investments across the AI ecosystem, anchoring core positions in OpenAI and Anthropic, according to Bank of America (BofA) estimates cited by Yahoo Finance. The commitments represent about 15% of the estimated $470 Bn in free cash flow Nvidia is projected to generate during 2026 and 2027.

But one of those investments could soon face a public test of what such holdings are worth. Anthropic investors expect the AI company to float at a valuation of $2 Tn or more in October, The Financial Times reported, citing half a dozen of its backers. Anthropic executives have not fixed an official valuation target; instead, investors arrived at the estimates through their own financial models.

That distinction matters because Anthropic has not stated it is seeking a $2 Tn valuation. The company confirmed in June that it had confidentially submitted a draft registration statement to the U.S. Securities and Exchange Commission for a proposed initial public offering, giving it the option to list after the regulator completes its review. The number of shares and offer price have not been set.

The private-market numbers are already substantial. Anthropic raised $65 Bn in May at a $965 Bn post-money valuation. The company said its run-rate revenue had crossed $47 Bn earlier that month as enterprise adoption of its Claude model grew. The round brought in venture firms, asset managers and sovereign wealth investors, including Singapore’s GIC and Temasek Holdings.

Investors speaking to the FT are betting on another sharp acceleration. They expect Anthropic’s annualised revenue to reach between $100 Bn and $120 Bn by the end of 2026, with that growth forming the basis for a valuation well above the company’s latest private-market mark. Because Anthropic lacks a directly comparable listed U.S. peer, investors have built their own models using valuations of companies viewed as major AI beneficiaries.

The Second Payoff

For Nvidia, the potential upside extends beyond selling the processors required to train and run large language models (LLMs). Its recent equity commitments include $30 Bn to OpenAI, up to $10 Bn to Anthropic and $5 Bn to Safe Superintelligence, according to BofA estimates.

The company’s chief executive, Jensen Huang, said in March that the latest OpenAI and Anthropic investments could be among the company’s last opportunities to back the two AI groups before they reach public markets.

That gives Nvidia dual exposure to the AI boom. The first remains its core business: supplying the processors, networking equipment and systems underpinning the build-out of AI infrastructure.

The second is financial, driven by stakes in the companies buying that computing capacity to develop new software, models and services.

The scale of the hardware business makes that investment strategy possible. Nvidia, which reports on a 52/53-week fiscal calendar rather than the standard calendar year, posted $81.6 Bn in revenue for its fiscal first quarter ended April 26, 2026, up 85% from a year earlier, while data centre revenue rose 92% to $75.2 Bn.

The company generated $48.6 Bn in free cash flow during the quarter and held $30.2 Bn in marketable equity securities alongside $43.4 Bn in non-marketable securities at quarter-end.

From Chips to Capital

Nvidia is now extending that financial role beyond its own balance sheet. On August 10, the company announced MoUs with Apollo Global Management, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to establish independent financing platforms intended to mobilise more than $500 Bn of third-party capital for AI infrastructure over time.

Nvidia said the proposed pools would help finance customers building AI computing infrastructure across its ecosystem. However, the arrangements remain subject to final agreements.

The combination marks a shift in Nvidia’s position in the AI economy. The company entered the boom primarily as a supplier of scarce computing power. It is increasingly an investor in companies buying that computing power and a participant in efforts to connect those customers with outside pools of institutional capital.

The Circular Logic

That model has drawn scrutiny. Bloomberg has mapped a web of transactions, widely described as ‘circular deals’, in which chipmakers invest in AI companies that are also major customers. These include arrangements involving Nvidia, Anthropic, OpenAI and rival chipmaker Advanced Micro Devices.

The concern is that when suppliers finance customers that subsequently spend heavily on their equipment, it becomes harder to separate underlying demand from the demand supported by the financing ecosystem around it.

The scale of third-party capital entering the sector offers a counterpoint. Nvidia’s proposed $500 Bn financing platforms would bring asset managers and other financial institutions into AI infrastructure funding, rather than leaving the chipmaker to support expansion solely through its own balance sheet.

But it also raises the stakes. More private capital is being mobilised around an industry whose largest companies are still establishing what their businesses will ultimately be worth in public markets.

The Litmus Test

Anthropic could become an early test of that value. Investor optimism persists despite mounting competition from Chinese AI developers, regulatory pressure and increasing customer sensitivity to the cost of using premium AI models. The company’s overall revenue growth slowed in June before rebounding, according to two investors cited by FT.

A listing would do more than provide liquidity to Anthropic’s early backers. Its $965 Bn private valuation was established through a funding round involving venture capital firms, asset managers, sovereign wealth funds and strategic investors. Public trading would expose those assumptions to a broader pool of investors and continuous price discovery.

For Nvidia, that could begin answering whether the extraordinary cash generated by its semiconductor dominance has created a secondary engine of value through its AI investment portfolio. If public investors support valuations close to those being modelled by Anthropic’s backers, Nvidia stands to participate in the AI boom both as an infrastructure supplier and a stakeholder in some of its fastest-growing client companies.

The exposure cuts both ways. Public markets unwilling to sustain the multiples established in private fundraising rounds would not erase Nvidia’s chip revenues. But they could expose a gap between the value private investors assign to frontier AI companies and what a broader market is prepared to pay.

After years in which private capital helped push AI valuations higher, Anthropic’s prospective IPO could begin answering a question that funding rounds alone cannot. How much of the wealth created on paper during the AI boom will survive public-market scrutiny and price discovery?

With inputs from agencies and The Financial Times