It’s beginning to look like Anthropic waited too late to go public. Third-quarter revenue numbers disclosed to investors on Wednesday implied that its growth has slowed from the second quarter, which we reported is likely to put a ceiling on its IPO valuation. Anthropic might have been better off going public in June, when its growth was still spectacular. (Of course, a June IPO might have been impractical given that that’s when SpaceX went public.)

As things stand now, Anthropic’s slowing growth may remind investors of a longstanding question surrounding the company: How much of a moat does it really have? Yes, its AI models are as good as they get, but increasingly, businesses don’t need really expensive frontier models. As we wrote earlier this week, developers are increasingly opting for other models that are good enough for most tasks. Anthropic’s aggressive pricing has prompted companies such as Microsoft to cut its spending on the former’s models, we reported 10 days ago. (This trend seems to be spreading).

This was not the day to be a publicly traded neocloud (as young AI data center firms are known). As the market sold off on bad inflation data—reinforcing certainty of another interest rate hike—neoclouds tanked. These companies typically are borrowing a lot to finance expansion. The worst hit in the stock sell-off was CoreWeave, whose stock fell 15.5%, fueled by chatter that its bonds were falling to worryingly low levels. 

CoreWeave CEO and co-founder Michael Intrator was defiant. He told his staff, according to a spokesperson: “When you are building a new type of technology company, there will be periods of time when the market misunderstands and acts irrationally.” Yeah, right. The issue may be more that Intrator doesn’t understand investors. He sold CoreWeave stock every week between April and the end of September, securities filings show, while his fellow co-founders have also been selling regularly. That can’t help but undermine investor confidence in the company. Intrator could win some brownie points by pledging to stop selling his stock for a while. 

It’s not just humanity that AI folks worry the new technology could wipe out—it’s publishers as well! (Publishers are, of course, people!) That’s what we learned from a court filing today by a group of news publishers including The New York Times Co. in a copyright infringement lawsuit against OpenAI and Microsoft. The filing cited an OpenAI executive writing—presumably in an internal memo—that publishers face “an existential threat” from AI products.

We knew the publishers worried that AI threatened their existence—The New York Times Co. made that point when it sued OpenAI and Microsoft back in 2023 (the case has since been consolidated with similar lawsuits filed by other publishers). What’s striking is that OpenAI and Microsoft internally acknowledge the threat—and recognize the implications for them as well. “Our AI content strategy has started a ‘doom loop’ that will hurt the performance of our models and the entire web at the same time. It is highly unusual that an end-product threatens the economic foundations of its essential suppliers,” according to a Microsoft document cited in the filing.

Put aside, for just a moment, all the arguments about AI safety. Today’s decision by the Federal Reserve to raise interest rates may be a bigger issue for the AI sector in the short term. Long-term bond yields had already risen, of course. But the Fed’s action—pushing up short- and medium-term borrowing costs—seems sure to crimp the debt-fueled AI boom, particularly given the prospects that another rate hike is possible later in the year. 

The most vulnerable to higher rates are surely small companies that have weak or no credit ratings but need to raise money, such as would-be neoclouds hoping to finance data center projects. Think Rum Group, which plans a big data center in Georgia for a deal with Anthropic but doesn’t have financing for the project yet. (For more on startups’ financing challenges, see here). But the ripple effects of higher interest rates will eventually affect everyone, including big tech firms spending a fortune to expand their data centers for AI.

When it comes to product names, companies tend to move in packs. In video streaming, almost every company included a plus sign in the name brands of their services (Disney+, Paramount+, Apple TV+—you get the picture). In tech, we’ve got Apple One, Google One and now Meta One (we used to have Amazon One, but that has been dropped). Does no one have any imagination? 

Aside from Amazon One, which was a palm-reader system for retail stores, all the other Ones are subscription services for various add-on offerings from the tech firms. Some of these have been around for years—Google One and Apple One have long offered extra amounts of digital storage, along with other perks—but AI is making subscriptions a far more important business model than it has ever been for tech. Google has added extra use of its AI service to its Google One offering. Meta One—unveiled on Tuesday—also offers extra AI usage, as well as some additional tools for using Meta’s social media apps. 

Forget all that chatter about AI killing us all. We’re back to looking at what really matters—the revenue! Oracle demonstrated on Thursday that demand for AI computing capacity remains pretty darn robust. Larry Ellison’s software and cloud firm reported 30% top-line growth for the quarter ending Aug. 31, a tad better than the company projected in June, sending its battered stock up 4.4% in after-hours trading.

The quarterly update demonstrated how much Oracle is benefiting from customers so desperate for capacity that they're paying up-front—what the company calls prepayments—for computing rental deals. Oracle brought in $11.4 billion in such customer prepayments in the quarter, doubling the amount of cash its operations generated. These payments help Oracle finance data center construction without dipping as much into its own pocket. While it spent $28.5 billion on capital expenditures in the August quarter, Oracle only burned $5 billion of its own money.

Oracle reported 30% growth in revenue of $19.3 billion for the quarter ending Aug. 31, a slightly better growth rate than the company projected in June, even as it burned $5 billion to expand the data centers that are driving its accelerating topline growth.

The software and cloud firm said ...

Italian digital conglomerate Bending Spoons is buying whiteboarding software firm Miro at a valuation of $1.355 billion, the latest in a series of purchases the Italian firm has done at knock-down prices.

The price for Miro is a huge discount to its peak valuation of $17.5 billion, dating ...

Are we sleepwalking our way into AI-caused extinction? It feels a little like that, given an Anthropic researcher’s X post on Tuesday night that there’s a greater than 10% chance that “AI could kill all humans” within the next decade! Despite that rather gruesome note, for tech executives posting on X today, it was business as usual. (See our detailed report on the episode here).

Plenty of dark humor circulated about the researcher’s post, however—this one joking about a Wall Street analyst asking Anthropic CEO Dario Amodei about the “extinction number” on an earnings call is possibly the best. That joke also highlighted the reality that Anthropic will shortly be marketing its IPO. How will its executives handle the warning on the Wall Street investor road show? 

Anthropic’s IPO paperwork is expected to become public sometime soon. But it was OpenAI finance chief Sarah Friar who drew a standing-room-only crowd Tuesday at San Francisco’s Palace Hotel, as Goldman Sachs kicked off its annual Communacopia + Technology conference with a Q&A with Friar. In her talk, Friar made clear how important price has become for determining model popularity—a potential problem for Anthropic, widely seen as the high-end AI provider.

Friar described how a price cut on OpenAI’s GPT 5.6 Luna model led to a tenfold increase in use, my colleague Anita Ramaswamy reported, giving OpenAI the highest market share on OpenRouter, a service developers use to access different models. We wrote in mid-August that Luna had generated more than twice the token usage of Anthropic’s Opus 5 and Sonnet 5 models on OpenRouter. Judging from Friar’s comments Tuesday, the increase in usage was enough to offset the price cut.

If at first people don’t pay attention to what you say, say it again. That appears to be Meta Platforms CEO Mark Zuckerberg’s philosophy. About two weeks after publishing a roughly 1,000-word opinion piece in The Wall Street Journal headlined “The AI Future Is for Everyone,” Zuckerberg published a 6,500 word version of the same piece on Meta.com on Monday. (Today’s essay carried a nearly-identifical headline, while the first half-dozen paragraphs were almost word for word what ran in the WSJ opinion piece.) Hint to Zuckerberg: Shorter is better.

Zuckerberg’s opus was part of a growing number of lengthy essays, published under the bylines of AI leaders, that discuss AI and society. Anthropic CEO Dario Amodei is a frequent contributor to the genre, with his most recent publication “Policy on the AI Exponential.” OpenAI CEO Sam Altman has his own blog, supplemented by weighty posts from his company like this one. Microsoft CEO Satya Nadella posts on X. For the most part, these essays are long and dry. In other words, they’re perfect for an AI chatbot to summarize. It makes you wonder if they’re meant to prove to consumers the value of AI!

Enterprise software firm Monday.com reported a 22% lift in second quarter revenue, but projected a slowdown in third quarter revenue growth to between 16% and 17%, partly because of the impact of a big layoff it undertook last month. Monday shares fell 9%.

At the same time, Monday.com said it ...

Intel is raising $15 billion in a public offering of its stock, taking advantage of the enormous improvement in investor sentiment it has enjoyed in the past year to raise money. Intel stock is up 175% over the past 12 months. Intel shares fell 4% Monday.

The iconic chip maker, which has ...

Meta Platforms CEO Mark Zuckerberg called for changes to U.S. policy to reduce the “friction” that American AI labs have to deal with in developing open-source models, compared to foreign counterparts, as a way of ensuring the U.S. leads in open-source AI. At the same time, Meta unveiled its ...

After last week’s earnings onslaught, things will likely quiet down this week. Aside from neoclouds CoreWeave and Nebius, no major tech earnings announcements are scheduled, which gives us a chance to assess what we learned in the past week. One big takeaway is that the enterprise software sector has bifurcated between those firms that are getting something out of AI and those that are not. 

It’s no secret Palantir is a winner—its revenue leapt 89% in the first half of this year. As we wrote, Palantir has succeeded in winning lots of business with AI agents and human consultants that help clients make AI work for them. Then there’s Shopify, whose software powers online storefronts for merchants. The company last week reported 34% revenue growth for the quarter, in line with the first quarter—but an improvement on last year’s growth rate, which in turn was faster than 2024’s. Shopify is benefiting from AI chatbots that are giving shoppers recommendations on what to buy, pulling in many cases from its catalog of merchant products. 

If you’re one of those executives at Alphabet, Meta Platforms or Amazon who deals with investors, here’s an idea: Put a large photo of SoftBank CEO Masayoshi Son on the wall of your office. Next time a visiting investor complains about your AI spending, point to the photo. That should end the discussion.

As much as big tech firms are spending on AI, at least they’re financing much, if not most, of it through internally generated cash. SoftBank in contrast has so far spent $54.6 billion on OpenAI shares by raising money through a combination of various kinds of loans and asset sales. And it’s now gone even further: On Thursday, SoftBank revealed that this week it had borrowed $10 billion using its OpenAI stake as collateral. As SoftBank has committed to put a further $10 billion into OpenAI in October, there’s a good chance it will make that payment using money borrowed against other OpenAI shares. 

Amazon has completed its $50 billion investment in OpenAI, putting in the remaining $35 billion in two stages in recent months, the commerce-and-cloud firm revealed in a securities filing on Friday.

Amazon agreed in late February to invest $50 billion into the ChatGPT creator, but it only put in ...

There’s no doubt about it: AI gave the big cloud firms a lift in business in the June quarter. We’ve already heard that from Google and Microsoft, and on Thursday, Amazon, the biggest cloud firm, followed suit by posting 37% growth in its Amazon Web Services cloud unit for the quarter. That’s 9 percentage points faster than it reported for the March quarter, and more than double the growth rate for the same period last year. 

Just as strikingly, AWS’ operating profit margin also rose, to 39.4%, the highest since the start of last year. To lift its operating margin while grappling with rising depreciation expenses on new chips and servers is an accomplishment. Amazon stock jumped 9% in response, suggesting investors were more focused on AWS’ accelerating growth than on the huge capital investments it is making to drive future growth. As a reminder of the size of those investments, Amazon burned $7.6 billion in cash in the quarter and raised its 2026 capital expenditures projection by $20 billion to $220 billion.