The banks that will take Anthropic public appear to be the same banks that are lending it money first. Anthropic is close to finalizing an expansion of its revolving credit facility to $15 billion, according to people familiar with the matter, with Morgan Stanley leading the arrangement and Goldman Sachs, JPMorgan Chase and Citigroup playing senior roles. Those same four banks have been reported as the principal underwriters for Anthropic’s expected initial public offering.
The order of operations is unusual and instructive. Companies typically borrow after they go public, using their listing to raise equity and establish a market valuation before asking banks for credit. Anthropic is doing the reverse: building a credit cushion while it is still private, and doing it through the institutions that stand to earn underwriting fees when the company lists. For the banks, the arrangement bundles two businesses, lending and equity issuance, into one relationship. For Anthropic, it means arriving at the IPO with a war chest already in place.
The size of the facility says something about what Anthropic plans to spend on. The company’s costs are dominated by computing, the accelerators and data-center capacity needed to train and run frontier models, and those costs run in the tens of billions of dollars a year across the industry. A $15 billion revolving credit line gives Anthropic the ability to commit to compute contracts on terms that assume the money will be there, and to negotiate those contracts from a position of strength rather than uncertainty.
The timing is the other notable feature. Reports have circulated that Anthropic could file its IPO prospectus as soon as next week, and the credit line is coming together in the same window. People familiar with the company’s thinking said the financing was planned in parallel with the IPO process rather than in response to it, a way of ensuring that whatever happens in the public markets, the company’s capital position does not become a constraint on its plans.
The move also sharpens the comparison investors will draw with OpenAI. Anthropic’s chief rival has raised enormous sums from Microsoft and others, and the two companies are widely seen as the leading pair in the race to build frontier models. Anthropic has said it intends to compete not just on model quality but on the scale of its commitments, and a $15 billion credit line is a direct statement about the company’s appetite for compute. The capital race between the two labs has been running for years; the credit line is the latest installment.
For the banks, the deal confirms that AI has become one of the most active arenas in investment banking. Morgan Stanley, Goldman Sachs, JPMorgan and Citigroup have all built practices around the sector, and they are competing for the mandates that come with the industry’s largest financing events. Anthropic’s IPO, whenever it arrives, is expected to be among the largest technology listings in years, and the credit facility is best understood as part of the courtship: the banks that lend now are positioning themselves for the underwriting fees later.
The structure of the facility, if it closes as reported, would put Anthropic in an enviable position. A revolving credit line is cheaper than equity and more flexible than a term loan; the company can draw on it when opportunities arise and repay when its cash flow allows. For a business whose capital needs are lumpy and tied to the timing of compute deliveries, that flexibility has real value, and the terms Anthropic has secured, with four of the world’s largest banks participating, suggest the company is being offered money on favorable conditions.
There is a risk embedded in the strategy, and it is the same risk that attaches to any buildout paid for with borrowed money: credit must eventually be repaid. Anthropic’s bet is that its revenue will grow fast enough to service a $15 billion facility, and that bet rests on the company’s ability to convert its models into a durable business. If the AI market softens before Anthropic’s revenue catches up with its commitments, the credit line that looks like strength today could become the measure of its vulnerability.
The prospect of a public listing changes the math in the company’s favor. Public companies can raise equity more easily than private ones, and an IPO would give Anthropic a currency, its own stock, that it could use for acquisitions, for employee compensation and for the kind of partnerships that have defined the AI industry. The credit line is best read as a bridge: it carries the company from its private capital structure to the day it can tap the public markets for whatever it needs next.
For now, the people familiar with the matter say the facility is close to done but not final, and the terms could still shift before the paperwork is signed. What is already clear is the direction of Anthropic’s strategy: borrow before you list, lock in the banks that will take you public, and arrive at the IPO with the balance sheet of a company that intends to keep spending. The credit line is a signal to the market about scale, about timing and about confidence, and it will be read as such when the prospectus finally appears.
