Getting your
Trinity Audioplayer ready...By Sridhar Natarajan, Katherine Doherty, Liana Baker and Shirin Ghaffary, Bloomberg
Bank of America Corp. handed a $520 million credit line to OpenAI in recent weeks after previously spurning the artificial intelligence giant’s request, according to people with knowledge of the matter.
The deal marks a notable pivot in how the second-largest US lender and its risk-wary chief executive officer, Brian Moynihan, approach money-losing AI startups that are driving global markets. As recently as last year, the bank’s executives were skeptical that ventures known for such voracious capital spending would be able to sustain their business models.
But OpenAI’s preparations for an initial public offering influenced the financial firm’s decision, said the people, who asked not to be identified describing internal deliberations. For a bank known for its “Thundering Herd” of Merrill Lynch wealth advisers, missing out on such a gigantic US stock listing would be seen as a black eye.
The lender’s commitment adds to an existing undrawn facility set up by competitors, boosting the capital available for OpenAI to more than $5 billion. The bank’s slow embrace of the first AI company to seize global attention reflects Moynihan’s distaste for leap-of-faith risk-taking for corporate clients, which he has drummed into his management team with a “responsible growth” mantra.
Bank of America slowly got comfortable financing AI and AI-adjacent companies once it felt certain that the market would support their business models, even if they weren’t profitable, one of the people said. Moynihan’s management team also took note of investors’ willingness to back such companies at valuations that weren’t tethered to their underlying earnings, adding to their level of confidence.
Representatives for Bank of America and OpenAI declined to comment.
OpenAI first tapped a syndicate of nine global banks in October 2024 when it arranged a $4 billion credit line. At the time, the AI venture said the “partnership with an exceptional group of financial institutions” would give it flexibility to invest in new initiatives.
Bank of America was noticeably absent from the lineup, which included all of its top US rivals, reflecting its more conservative approach.
The firm would later wade into providing financing for other parts of the AI ecosystem, such as Oracle Corp.’s data center buildout in Michigan. But even then, the lender remained on the sidelines again as OpenAI increased the revolving credit line to $4.7 billion in March, adding two other banks.
Bank of America still needed more time to conduct due diligence and push the credit decision through its internal controls.
Battle Scars
Moynihan, 66, who took the helm in the aftermath of the 2008 financial crisis, is widely credited for leading Bank of America out of disastrous decisions made by his predecessors.
But investors have been less impressed in recent years as he has sought to position his firm as the safest bet in any future downturn.
His philosophy aims to avoid blowups, even if that sometimes means missing opportunities to maximize profits. It’s a lesson ingrained in his management style from his earliest days on Wall Street. More than two decades ago, when he was still a rising executive at FleetBoston, he watched his bank suffer big losses in Argentina.
In 2017, when Wall Street firms took hits on margin loans tied to the collapse of South African retailer Steinhoff, most shrugged it off. But under Moynihan, Bank of America curtailed that line of business just as financing clients became a major revenue driver for its peers.
A few years later, the firm made a costly decision to plow hundreds of billions of dollars into long-dated, low-yielding securities during the rock-bottom interest rates that prevailed during the pandemic. While the firm now says it’s actively replacing those securities with other assets that can help it earn more money, the blunder eroded earnings for years and left the bank hamstrung in playing offense.
The result: Despite hitting a new high this week, Bank of America’s stock has significantly underperformed all of its peers in the past five years. And middle managers have privately bemoaned the opportunities they missed because they couldn’t take certain risks.
IPO Timing
Rising Silicon Valley giants have come to expect banks to support them with credit lines as they scale up and prepare for eventual stock debuts. Capital-markets bankers vying for high-profile IPO roles often expect their bosses to provide those loans to help their franchises compete for deals, even if such quid-pro-quos aren’t explicitly spelled out.
For its part, the ChatGPT creator worked with Goldman Sachs Group Inc. and Morgan Stanley when it submitted a confidential regulatory filing for a potential IPO, and it has held discussions with other banks to add them to the roster of advisers on the listing. The AI venture, led by Sam Altman, has been considering how soon to go public and may wait until next year, potentially following rival Anthropic PBC. Both companies were initially racing to make their Wall Street debuts as soon as this year to secure fresh capital and support their heavy spending — but private fundraisings and credit lines give them more leeway.
OpenAI already raised capital at an eye-popping $852 billion valuation this year, demonstrating the willingness of investors to keep supporting the no-costs-spared race to dominate the AI landscape.
And when the company disclosed that it had filed its confidential IPO paperwork, it added a note of caution.
“We have not decided on timing yet; it may be a while because there are things we want to do that are likely easier as a private company,” OpenAI said. “It’s a complicated set of tradeoffs and this gives us the option to go public sooner if that ends up being best.”
–With assistance from Peter Eichenbaum.
More stories like this are available on bloomberg.com ©2026 Bloomberg L.P.