Salesforce CEO Marc Benioff at the World Economic Forum in Davos, Switzerland, on January 23, 2025 (Halil Sagirkaya/Getty Images)

Eye on AI

Salesforce beats on earnings but posts light full-year guidance

The SaaS company posted results for the fourth quarter and full 2026 fiscal year on Wednesday.

Rani Molla

Salesforce reported fourth-quarter earnings for fiscal year 2026 after the bell Wednesday, beating analysts’ earnings expectations but providing light full-year 2027 guidance. The company posted:

Adjusted earnings per share of $3.81, beating Wall Street’s expectation of $3.05.
Revenue of $11.2 billion, in line with the FactSet consensus estimate of $11.19 billion.
Revenue guidance for fiscal year 2027 of $45.80 billion to $46.20 billion and adjusted EPS of $13.11 to $13.19, which is potentially lower than the $46.1 billion in revenue and $13.12 EPS analysts had expected.

The stock is down over 4% after-hours.

Reflecting its “strong trajectory,” Salesforce authorized a $50 billion share buyback and increased its dividend by nearly 6% to $0.44 per share.

“Salesforce delivered a record Q4 as our customers’ shift to the Agentic Enterprise surges, fueling NNAOV acceleration in H2 FY26,” Robin Washington, Salesforce president and chief financial officer, said in the press release.

Fourth-quarter GAAP operating margins fell to 16.7% from 18.2% a year earlier, in part due to higher stock-based compensation and increased amortization of acquisition-related costs.

On the earnings call today, investors will be looking for more information on adoption of its AI offerings, which have been suffering from narrative headwinds.

After rallying last quarter, Salesforce shares have fallen about 30% year to date as investors weigh how advances in AI — including new enterprise-focused features from companies like Anthropic — could threaten parts of its core business. A number of analysts have cut price targets amid these concerns.

Still, some bulls say the sell-off is overdone. Wedbush Securities analyst Dan Ives has called it a buying opportunity, arguing that entrenched incumbents like Salesforce are well positioned to monetize AI among their existing user base.

Recent commentary from Anthropic’s own AI event suggests a more nuanced outcome. As Deutsche Bank wrote afterward, “We have even greater conviction that model providers are unlikely to displace software incumbents and are instead positioning themselves and their agents to be an orchestration layer on top of existing and incumbent systems.”

Rani Molla5h

Amazon, Google, Meta, Microsoft, Oracle will agree to “build, bring, or buy” AI data center power

A month after President Trump called on Big Tech companies to “pay their own way” for data center energy — and a day after Trump pledged as much in his State of the Union address — a number of tech’s biggest companies are planning to make it official, according to a report from Fox News.

Alphabet, Amazon, Meta, Microsoft, and Oracle, in addition to OpenAI and xAI, plan to sign agreements at a March 4 White House event committing to supply their own electricity for new AI data centers.

“Under this bold initiative, these massive companies will build, bring, or buy their own power supply for new AI data centers, ensuring that Americans’ electricity bills will not increase as demand grows,” White House spokeswoman Taylor Rogers told Fox.

Already, Amazon, Microsoft, and Meta have committed to as much in recent data center announcements.

Scoop: Trump brings Big Tech to White House to curb power costs amid AI boom

Already, Amazon, Microsoft, and Meta have committed to as much in recent data center announcements.

Jon Keegan

Anthropic doesn’t want Claude controlling autonomous weapons. The Pentagon may not give them a choice.

A looming Friday deadline to get onboard with the Pentagon’s demands could lead to an unraveling of years of the company’s work with the government.

Rani Molla

Waymo announces Chicago and Charlotte expansion as driverless rollout accelerates

The company is currently operating in 10 US cities.

The Streets of San Francisco, the Waymo way

Rani Molla11h

Uber to roll out robotaxis with Wayve in 10 markets

Uber also has its sights set on 10 robotaxi markets — a milestone that Alphabet subsidiary Waymo reached yesterday.

As part of its latest $1.5 billion funding round, autonomous tech startup Wayve announced that it will be helping to power robotaxis on Uber’s network, with its first launch in London this year. That’s followed by “plans to scale to more than 10 markets globally.”

The companies didn’t specify the vehicle model, but said Wayve’s AI Driver will be deployed in L4-capable vehicles from participating automakers. Uber will own and operate the fleet.

Microsoft, Nvidia, Mercedes-Benz, Nissan, and Stellantis were also part of the fundraise.

Separately, an Uber filing Tuesday showed that newly appointed CFO Balaji Krishnamurthy purchased $1.6 million in company stock.

Uber shares are up about 1% premarket.

Wayve secures $1.5B to deploy its global autonomy platform

The companies didn’t specify the vehicle model, but said Wayve’s AI Driver will be deployed in L4-capable vehicles from participating automakers. Uber will own and operate the fleet.

Microsoft, Nvidia, Mercedes-Benz, Nissan, and Stellantis were also part of the fundraise.

Separately, an Uber filing Tuesday showed that newly appointed CFO Balaji Krishnamurthy purchased $1.6 million in company stock.

Uber shares are up about 1% premarket.

Hyunsoo Rim12h

The way different states use ChatGPT could tell us a lot about their economies

Last Friday, OpenAI launched a new public dashboard tracking global non-enterprise messages on ChatGPT sent between the summer of 2024 and the end of 2025.

Among the 118 countries analyzed, the US ranked 25th by number of messages sent per capita, with a little over three-quarters of all conversations clustering around just three subjects: practical guidance (29%), writing (27%), and seeking information (19.6%). Zooming into the state level, it becomes clear that not all regions are using the chatbot in quite the same way.

DC topped the list of ChatGPT messages per capita — echoing patterns seen in the use of Anthropic’s Claude — and it’s not hard to see why. The capital, dense with federal agencies, think tanks, and law firms, runs on drafting documents, from memos and policy briefs to endless email chains. It makes sense, then, that nearly a third (32%) of DC users lean on AI for writing, above the national average of 27%. The pattern holds across other top-ranked states, too: New York (No. 3), California (No. 4), and Washington (No. 8), for instance, all show writing as their top use case.

Flip to the other end of the rankings, where the states that use ChatGPT the least sit, and the picture looks different. In West Virginia (No. 51), South Dakota (No. 48), Mississippi (No. 46), and Arkansas (No. 45), practical guidance is the dominant use case, often accounting for 33% to 35% of all prompts.