Demonstrators protest outside the courthouse as jury selection begins in the lawsuit between Elon Musk and OpenAI on April 27, 2026 in Oakland, Calif.Benjamin Fanjoy—Getty Images

A federal judge seated a nine-person jury on Monday as the legal battle between AI friends-turned-foes Elon Musk and Sam Altman begins.

Tesla, SpaceX, and DOGE aside—plus that whole “world’s richest man” thing—Musk is a cofounder of OpenAI. Altman is, too, as well as its current CEO. 

Musk sued Altman and the company (plus president Greg Brockman for good measure) in 2024 alleging that they didn’t stick to their commitments to keep OpenAI a nonprofit. Musk, of course, now directly competes with OpenAI through xAI, which merged with SpaceX earlier this year.

Musk is trying a number of tactics to inflict pain on his old comrades. He seeks to have Altman and Brockman removed from their posts. He claims $150 billion in damages. He is pressing to have OpenAI fully returned to nonprofit status, rather than the for-profit-in-a-nonprofit-trench-coat structure the company adopted to justify its unprecedented fundraising. 

The judge for the case has divided the trial in two—one portion for liability, the other for remedies. The jury will only give an advisory verdict on the liability portion, which is expected to wrap up in four weeks. —AN

Turns out the outrage over misuse of AI isn’t limited to Palantir employees.

More than 600 Google staffers—including some from its DeepMind unit—have demanded that the company reject a deal with the U.S. military that would allow Google’s Gemini AI models to be deployed in classified military operations.

“We want to see AI benefit humanity; not to see it being used in inhumane or extremely harmful ways,” the employees wrote in the petition. “This includes lethal autonomous weapons and mass surveillance but extends beyond.” 

The letter was addressed to CEO Sundar Pichai and signed by directors and VPs, among others. 

According to the Washington Post, Google proposed contract language to prevent Gemini’s use in domestic mass surveillance or autonomous weapons without humans “in the loop,” as it’s known. But the Pentagon much prefers that language to say “all lawful uses,” according to the report.

It’s not the first time Google employees have pushed for a change in the company’s government dealings. In 2018, staffers succeeded in pushing Google to abandon a Pentagon program that would have integrated its AI into drone operations.

But times have changed. Today, Google and its rivals seem more keen to secure business from the public sector—and perhaps justify their extraordinary AI spending. —AN

Just when it seemed like OpenAI and its largest outside shareholder, Microsoft, had worked it all out…things have changed again. 

The AI company announced on Monday a new agreement between the firms that lets OpenAI sell its wares on all cloud computing platforms, not just Microsoft’s Azure. The new terms allow OpenAI to secure more compute as it scales its enterprise services to compete with rival Anthropic; both companies are plowing toward IPOs. 

(OpenAI’s current valuation: $852 billion. Anthropic’s: $380 billion, though secondary markets have recently pushed them to be neck-and-neck….at $1 trillion. Yeesh.)

So what’s in it for Microsoft? Quite a bit. Redmond will remain OpenAI’s primary cloud partner (with a promise to use $250 billion in Azure services by 2032) and enjoys a license to OpenAI’s intellectual property through 2032. It will retain the right to make OpenAI products available first on Azure and lose the obligation to pay OpenAI a share of Microsoft’s revenue for offering OpenAI models on Azure. 

The tech giant will also receive a guaranteed—and that’s an important word—20% of OpenAI revenue until 2030. (“Though the total will now be subject to an undisclosed cap,” Reuters notes.) What’s more, that pesky line about OpenAI halting payments to Microsoft if it achieves artificial general intelligence, a.k.a. AGI, has been removed.

One gets flexibility, the other, certainty. A win-win? So far, so good. —AN

—Rural U.S. communities to AI data centers: Drop dead!

—Last month was the worst for reported tech job layoffs in at least two years.

—Amazon Now, the quick commerce service, will expand to 100 Indian cities.

—Meta will reportedly unwind its Manus acquisition after China rejected it.

—Lightelligence jumps 400% in Hong Kong debut. The Shanghai firm makes photonics chips.

—Starboard takes a major stake in Dynatrace. The activist investor’s moves sent shares up by 7%.

—35% of new websites published since ChatGPT’s launch were AI-generated to some degree, according to new analysis.