Thanks for reading The Briefing, our nightly column where we break down the day’s news. If you like what you see, I encourage you to subscribe to our reporting here.
Greetings!
Oracle managed to calm some frayed nerves on Wall Street on Tuesday. The cloud and software firm, which has become a focal point for investors’ anxiety about AI-related spending, reported slightly better than projected revenue for its February quarter. More importantly, for the second time in two quarters, it raised its revenue projection for next fiscal year, which starts in June. Oracle now projects revenue in fiscal 2027 to rise 34%, more than twice this year’s likely growth and four times its average top-line growth since 2021. Its battered shares rose 9% in after-hours trading.
What has stressed out investors—and cut the stock in half over the past six months—is that Oracle has borrowed tens of billions to build out data centers for customers such as OpenAI whose ability to pay for services a few years into the future likely rests on future fundraising. Oracle tried to assure investors all was fine: It pointed out that “some of the largest consumers of AI cloud capacity have recently strengthened their financial positions quite substantially,” a reference to its recently announced $110 billion fundraising. OpenAI doesn’t get that money all at once, however. And one of the investors in the funding round, SoftBank, is itself trying to borrow the money it needs for its OpenAI investment, according to Bloomberg. The AI build-out is a bit of a financing chain.
Oracle executives, on a call with analysts, also emphasized how the company has reduced the funding it needs to raise for future data center deals by arranging for customers to pay up front or bring their own AI chips. What Oracle executives haven’t explained, however, is how much those deals have cost it in revenue. It seems unlikely customers would pay the same amount if they’re bringing their own chips or paying for stuff in advance.
Oracle also hasn’t yet specified how the shift to these kinds of deals will affect its overall capital expenditures on data centers. Today it projected that capex for the full 2026 fiscal year, ending in May, would be $50 billion—more than twice the capex for the previous fiscal year. As a result, analysts estimate Oracle will burn $23 billion this fiscal year, according to S&P Global Market Intelligence. That’s a lot when you consider that Oracle produced a total of $25 billion in cash in the three years between 2022 and 2024. When asked for the fiscal 2027 capex outlook, executives punted. Investors might have come out of today’s earnings reassured, but the risks remain.
Benioff Bonds
Salesforce is getting bold. On the same day Amazon went into the bond market to raise as much as $50 billion for its AI data center expansion, Bloomberg reported that Salesforce was planning to sell between $20 billion and $25 billion of bonds to fund a $50 billion stock buyback. That’s a ton of debt for Salesforce, which had just $14 billion of debt outstanding as of Jan. 31, netting $9.5 billion of cash.
We should have seen this coming. CEO Marc Benioff dropped hints on the last earnings call, when he said Salesforce was “not using debt effectively.” He talked about the company’s stock price—shares have dropped 26% so far this year—and the fact that Salesforce has issued shares in a couple of big acquisitions over the years. “Now is the opportunity to take some of that stock back out of the market…and I think debt is a great way to do that,” he said.
With $50 billion, Salesforce could buy back 250 million shares at its current stock price. That would shrink its outstanding shares by about a quarter, reducing the share count to 673 million, the lowest number in about a decade.
Still, credit ratings agencies weren’t thrilled. Moody’s, for instance, downgraded the company’s rating to A2 from A1, still in the middle of the investment-grade scale, citing what it called a “material shift in financial policy including a higher tolerance for debt.” Salesforce stock, meanwhile, dipped roughly 2% to around $195, in the same ballpark where the stock has traded for the past month. A buyback of this size is designed to juice the stock, but it didn’t seem to impress equity investors, either.
In Other News
• Nvidia and Thinking Machines Lab, the model developer co-founded by former OpenAI Chief Technology Officer Mira Murati, announced a strategic partnership on Tuesday to deploy at least 1 gigawatt of servers powered by Nvidia’s forthcoming Vera Rubin chips, according to a blog post. Nvidia has also made a “significant investment” in Thinking Machines.
• Google is providing its Gemini Agent Designer to the Department of Defense for unclassified work, the company announced on Tuesday.
• Former Meta Platforms chief AI scientist Yann LeCun has raised more than $1 billion for his new AI startup, Advanced Machine Intelligence Labs, from investors that included Nvidia, Greycroft, former Google CEO Eric Schmidt and Jeff Bezos.
• Meta Platforms has hired Matt Schlicht and Ben Parr, the creators of Moltbook, a social network designed for AI agents. The company said they will join its AI division, Meta Superintelligence Labs. Meta also bought Moltbook itself.
• A judge ordered Perplexity to temporarily stop accessing Amazon’s website, a victory in the e-commerce giant’s lawsuit against the AI search startup. Amazon had sued Perplexity in November, arguing that it was fraudulently accessing the Amazon site through the Perplexity AI browser, Comet, in order to make purchases on behalf of its users.
Today on The Information’s TITV
Check today's episode of TITV in which we unpack Nvidia‘s deal with Thinking Machines Lab, OpenAI’s latest acquisition and Meta bringing on Moltbook's founders.
Recommended Newsletter
Start your day with Applied AI, the newsletter from The Information that uncovers how leading businesses are leveraging AI to automate tasks across the board. Subscribe now for free to get it delivered straight to your inbox twice a week.
Join The Information at the New York Stock Exchange on Monday, April 27, to hear from top executives and investors on how the rapid buildout of AI is reshaping tech, finance, and capital markets
