In one of the largest financings tied to the artificial intelligence boom, Bank of America, Blackstone, Citi, and Morgan Stanley are assembling $63 billion to purchase AI chips and networking equipment from Broadcom and lease them to Anthropic, the startup behind the Claude chatbot, according to Reuters.
The deal is structured through a special purpose vehicle (SPV), a standalone company created solely to hold the hardware and collect lease payments from Anthropic. Deliveries are expected to begin in 2027 under a previously announced compute-capacity agreement between Anthropic and Broadcom.
Bank of America, Citi, and Morgan Stanley are arranging the "first-lien" debt, meaning these lenders would be first in line to recover their money if the SPV were forced to seize assets. Blackstone, a major alternative asset manager, is also participating, though the exact split of equity and debt among the parties has not been disclosed.
Why this deal matters
Anthropic, like other leading AI labs, needs enormous computing power to train and run its models. Buying that hardware outright would require massive upfront spending, which is why leasing arrangements are becoming more common. By using an SPV, the banks and Blackstone can fund the purchase while Anthropic pays rent over time, preserving its cash for other uses.
This structure also spreads risk. If Anthropic fails to make payments, the lenders have a claim on the physical equipment, which retains value. The involvement of major banks and Blackstone signals confidence in both Anthropic's ability to pay and the long-term demand for AI infrastructure.
The deal comes as Anthropic is reportedly preparing for an initial public offering (IPO). The company has filed for a Nasdaq listing, and its IPO filing revealed soaring revenue alongside mounting losses, a common pattern for high-growth tech firms. A successful listing could be one of the largest in recent years, and this financing deal helps secure the computing capacity Anthropic needs to keep growing.
What it means for investors
For everyday investors, this deal is a reminder of how deeply financial institutions are betting on AI. Banks and asset managers are not just lending to AI companies; they are creating complex financing vehicles to fund the physical infrastructure that powers them. This trend is also visible in other sectors, such as Apollo backing SpaceX AI data centers.
If you own shares in Broadcom, this deal could be a positive sign, as it locks in future demand for its chips. For those invested in the banks involved, it represents a new revenue stream, though it also carries credit risk. For Anthropic's future public shareholders, the lease payments will be a recurring cost that could weigh on profitability, even as revenue grows.
The scale of the financing—$63 billion—underscores the enormous capital requirements of the AI industry. It also highlights the growing role of private credit and alternative asset managers like Blackstone in funding major technology projects, a shift that has been putting AI spending back in focus for investors.
Risks and watchpoints
While the deal is structured to protect lenders, there are risks. If AI demand slows or Anthropic's growth falters, the value of the underlying equipment could decline, and lease payments could be at risk. The first-lien structure gives banks priority, but it does not eliminate the possibility of losses.
Investors should also watch how this deal affects Anthropic's balance sheet ahead of its IPO. The company's IPO filing has put AI back in the spotlight, and the terms of this lease will be scrutinized by analysts and potential shareholders.
For now, the deal is a strong signal that major financial players believe AI infrastructure will remain in high demand for years to come. As AI stocks have rallied on the back of Anthropic's IPO plans, this financing arrangement adds another layer of confidence—and complexity—to the AI investment story.
