Meta has reportedly notified thousands of employees that they’ve been laid off as the company attempts to compensate for its hefty AI investments. In an email from Meta management shared by Business Insider, impacted staffers were told that the planned headcount reduction was part of the company’s “continued effort to run the company more efficiently and to allow us to offset the other investments we’re making.”

Meta lays off thousands of employees to offset AI investments

8,000 staffers are reportedly affected, around 10 percent of the company.

8,000 staffers are reportedly affected, around 10 percent of the company.

Reports of an upcoming wave of layoffs started circulating in March, though at that time Meta was believed to be cutting up to 20 percent of its total company headcount. According to a recent memo shared in May, the layoffs are now believed to be impacting approximately 8,000 people, around 10 percent of Meta’s 78,000 employees.

The layoffs follow Meta forecasting in January that it will spend $115 billion–$135 billion in capital expenditures in 2026, which would be used to “support our Meta Superintelligence Labs efforts and core business.” That’s almost double the $72.22 billion spent by the company in 2025 by comparison. Alongside cutting active roles, Business Insider reports that Meta is moving more than 7,000 staffers to work on new AI initiatives, and 6,000 open roles are being closed, according to Bloomberg.

“We want to say again that we’re grateful for your contributions. Your impact at Meta has been an important part of our story,” Meta said in closing the memo to laid off employees.

Some of those impacted have posted about being laid off on LinkedIn, showing their Meta employee badges and confirming that the cuts are now underway. One former employee mentions she was let go alongside “8,000 metamates.” We have reached out to Meta to confirm how many employees were affected.

Read on theverge.com →

Sam Altman and Elon Musk are facing off in a high-stakes trial that could alter the future of OpenAI and its most well-known product, ChatGPT. In 2024, Musk filed a lawsuit accusing OpenAI of abandoning its founding mission of developing AI to benefit humanity and shifting focus to boosting profits instead.

After nearly a month, with the trial featuring testimony from Musk, Altman, Microsoft CEO Satya Nadella, OpenAI cofounder Greg Brockman, former OpenAI board member and mother of several of Musk’s children Shivon Zilis, and a few others, the jury deliberated for a couple of hours before returning to the “room full of untrustworthy, unreliable people all fighting with each other” with a verdict, deciding to dismiss all charges due to the statute of limitations.

Musk was a cofounder of OpenAI and claims that Altman and Brockman tricked him into giving the company money, only to turn their backs on their original goal. However, OpenAI claimed that “This lawsuit has always been a baseless and jealous bid to derail a competitor” in a bid to boost Musk’s own SpaceX / xAI / X companies that have launched Grok as a competitor to ChatGPT.

In his lawsuit, Musk asked for the removal of Altman and Brockman, and for OpenAI to stop operating as a public benefit corporation.

People to Know

Plaintiff

Elon Musk — plaintiff, OpenAI cofounder and now CEO of rival xAI

Steven Molo — lead counsel for the plaintiff

Jared Birchall — manager of Musk’s family office

Shivon Zilis — former OpenAI board member who shares multiple children with Musk

Defendant

Sam Altman — defendant, CEO of OpenAI

William Savitt — lead counsel for the defendant

Greg Brockman — president of OpenAI as well as a cofounder

Ilya Sutskever — former chief scientist at OpenAI and a cofounder

Judge

Yvonne Gonzalez Rogers — aka YGR, trial judge

Here’s all the latest on the trial between Musk and Altman:

Musk v. Altman: Much ado about nothing

Today I’m talking with Liz Lopatto, who spent the last month covering the Musk v. Altman trial in all its chaos. You’ll hear her describe the courthouse as a “zoo” and explain that there were protests of one kind or another happening outside every day.

Both Elon Musk and Sam Altman are big personalities, and people have a lot of feelings about both of them and the AI industry. And in the end… nothing happened! The jury found that Elon had filed his lawsuit after the statute of limitations had run out. You’ll hear Liz explain exactly what’s going on there.

Read Article >Musk v. Altman proved that AI is led by the wrong people

The tech trial of the year, Musk v. Altman, was ultimately a fight for control. Elon Musk argued that Sam Altman, with whom he helped found the now-massive company OpenAI, shouldn’t direct the future of AI. Altman’s lawyers, in turn, poked at Musk’s own credibility. A jury came to a verdict on Monday after just two hours of deliberation, dismissing Musk’s claims due to the statute of limitations.

In a strictly legal sense, three weeks of testimony added up to nothing. But the trial offered a more damning broader takeaway: Almost nobody in this saga seems worth trusting. Some of the most powerful people in tech seem temperamentally incapable of dealing with each other honestly. And if that’s true, it raises a bigger question: Why are they in control of a trillion-dollar industry that’s set to upend people’s lives?

Read Article >Elon Musk loses his case against Sam Altman

After around two hours of deliberation, the jury has reached a unanimous verdict in Musk v. Altman, the tech trial of the year. The group found that two claims were barred by the statute of limitations, and a third failed thanks to the dismissal of one of these.

The jury here is an advisory jury, meaning the group is installed solely to offer another opinion to the judge, and its verdict is technically not legally binding. Ultimately, US District Judge Yvonne Gonzalez Rogers is the ultimate legal authority — and she accepted the decision.

Read Article >- The jury has delivered a unanimous verdict.

That was quick (about two hours).

Elon Musk lost his case against Sam Altman

Hayden Field - An observer has just been ejected from the court by the US marshals.

I assume because he was recording, since the marshal said, “Give me your phone.” There have been several incidents of people attempting to record or take pictures throughout the trial — but I honestly am not sure why you’d record today of all days.

- C. Paul Wazzan is the expert called by Musk to determine damages.

Unfortunately he does not have a lot of details YGR is asking for. He doesn’t know how many investments Musk has made (11 to date, according to Pitchbook), or how many were successful. He’s getting some pretty tough questioning from YGR in the direct exam. Among the things she’d asked, he didn’t know how many startups fail in Silicon Valley, or the success rate for assorted VC firms.

Musk v. Altman accomplished nothing but airing dirty laundry

Today was closing arguments in the Musk v. Altman trial, and I almost feel bad writing about the unbelievable demolition derby I just witnessed. Steven Molo, Musk’s lawyer, stumbled over his words. He at one point called Greg Brockman — a co-defendant — Greg Altman. He erroneously claimed that Musk wasn’t asking for money and had to be corrected by the judge. He made it clear we’ve heard from many liars over the past few weeks, but offered little evidence for Musk’s actual legal claims.

OpenAI’s lawyer Sarah Eddy countered this by simply arranging the mountain of evidence that the company introduced in chronological order. She didn’t spend time trying to pretend anyone in this trial is especially reliable. She did, however, get the zinger of the day, about Musk: “Even the mother of his children can’t back his story.” William Savitt, who took the defendant baton after her presentation, demonstrated the number of times Musk “didn’t recall” some critical detail — and wondered how a sophisticated businessman couldn’t understand or read a four-page term sheet OpenAI had sent to him.

Read Article >- “I told you in my opening statement you wouldn’t hear very much from Microsoft, and you haven’t.”

God bless. We are in the Microsoft closing statements. “Microsoft never found a single page of a single document” that referenced Musk’s alleged restrictions on his donations during the due diligence process.

Microsoft doesn’t want any of this

Elizabeth Lopatto - “I feel like I’m going to miss you all,” Savitt tells the jury.

He may be laying it on thick, but he did get a big laugh.

- Savitt calls out the fact that Musk is abroad with President Trump today.

He reminded the jury that Musk isn’t in the courtroom while Altman and Brockman are. (Musk posted yesterday that he was en route to Beijing on Air Force One.) “They are here because they care a lot about this,” Savitt said. “Mr. Musk isn’t here. Mr. Musk came to this court for exactly one witness — Elon Musk — and he hasn’t been seen since. Now he’s in parts unknown.”

- Savitt says Musk has “selective amnesia.”

“He claims to have heard things high atop a windy hill where no one else can hear,” Savitt told the jury. (Strange phrasing, but after the bridge metaphor from Molo, I wouldn’t expect anything less.) He also says Musk has “unclean hands” due to his “unconscionable conduct” related to the claims he’s bringing. “Only after OpenAI succeeded, against Musk’s prediction, only then did he start threatening litigation,” Savitt said.

Behold, the Elon Musk jackass trophy

Yesterday, in Musk v. Altman, before the jurors came in, Sam Altman’s team passed up what looked — from a distance — like a Little League trophy. It was not. Judge Yvonne Gonzalez Rogers had the lawyers read the inscription aloud for the press: “Never stop being a jackass.” It’s a commemoration OpenAI employees bought for research scientist Josh Achiam, who testified yesterday.

How exactly did this come up in a trial about nonprofit contract law? Allegedly, when Elon Musk was leaving OpenAI, he talked about wanting to race ahead of Google. Achiam, who worked on AI safety, asked if that was really such a good idea. Musk called him a jackass. Years later, Musk is portraying his lawsuit as an attempt to avoid AI causing serious harm — something that, Altman’s team suggests, wasn’t high on his list of concerns back then.

Read Article >- Savitt is talking about the statute of limitations.

He said that by his calculations, people said things like “I don’t remember” and “I don’t recall” between 150 and 200 times during this trial so far — using this to bolster his argument that Musk had waited too long to bring his claims.

- Here’s the jackass trophy that the jury didn’t get to see.

You may remember that yesterday I was completely tickled by the possibility that the jury might get to see this. Even YGR seemed tickled by it. Unfortunately, she ruled that discussing it was fine but unless the Musk team gave them reason to introduce it, the jurors wouldn’t see so much as a photo. But this is the trophy Josh Achiam got for getting yelled at by Elon Musk.

- We are back from our break. William Savitt is taking it home for the OpenAI defense.

There’s one more thing that Savitt is harping on. “Has the OpenAI nonprofit respected its general founding principles?” The question doesn’t matter, legally, since Musk didn’t create a charitable trust, but Savitt is going to spend some time on this because Molo emphasized it.

- Tesla AI is evidence of Musk’s failure, Eddy says.

Musk doesn’t want to admit that trying to build an AGI lab in Tesla was a failure — whether that was by acquiring OpenAI or trying to poach all its talent, maybe even putting Altman on the board. Eddy suggests this case is revenge on OpenAI for succeeding.

- “The documents tell the truth here,” Eddy says.

That’s kind of where I’ve landed! The idea of the “adjunct” for-profit (Eddy says this is a moving target, and though Musk used it twice in testimony, when Savitt used it, Molo objected and accused him of making up a term) doesn’t show up in any of the brainstorming structure documents. We do see parallel for-profits, and the idea of a conversion to a for-profit and shutting down the nonprofit. Jared Birchall also testified that he filed to register a company for this.

- Eddy suggests that Musk’s donated Teslas were, effectively, bribes.

They went to Ilya Sutskever and Greg Brockman, among others, right before he proposed that he get 62.5 percent of a for-profit company. We are now looking at tax forms and letters — neither of which show any specific purpose. Jared Birchall also testified that there was no specific purpose for the donations. Shivon Zilis doesn’t remember it. Sam Teller doesn’t remember it. This is like watching the Warriors play a team of 6-year-olds.

- Eddy is, wisely, leaning hard on chronology in explaining their defense.

Chronology and documents. Musk’s performance on the stand does give credence to the suggestion from Eddy that Musk “took his marbles and went home” when he couldn’t get his way.

- Sarah Eddy is giving the closing argument for OpenAI.

She opens with a banger. Musk has said he made donations with strings attached. “Even the mother of his children can’t back his story.”

- Molo is done presenting Musk’s closing argument.

I do wonder how this would have played in the hands of a better lawyer. Molo’s book report did not overwhelm me with confidence in his case, particularly because a lot of his point-blank assertions were profoundly arguable.

- For all of the very irritating testimony about “the blip,” Molo hasn’t convincingly connected it to his case.

In all that chaos, Microsoft did suggest board members. But OpenAI didn’t take those suggestions — except one, well after the crisis. I don’t know man, I don’t really understand how this goes to the Microsoft case. It might be a suggestion the nonprofit board doesn’t really control the for-profit — but Helen Toner’s and Tasha McCauley kind of came off as amateurs in their approach, not least because there was no investigation before the firing.

- Molo is now attempting to make a case against Microsoft.

So far he’s got “they’re a for-profit corporation,” “they know Musk was a co-founder,” and “they read the announcement OpenAI existed.” This is easily the thinnest part of a very thin case — on the OpenAI part, there’s at least Brockman’s diaries.

- During our break, the jurors were out of the room, and the lawyers were beefing again.

During his speech, Molo told the jury he wasn’t asking for money. That is in fact not true — otherwise I wouldn’t be sitting through phase 2 of the trial next week. “You slipped it in nicely,” YGR says. But Molo needs to retract that statement.

- We are back on the millions and billions of OpenAI equity that employees have interest in.

I am just going to break from telling you what Molo is saying to say what my personal impression was from sitting here all these weeks: Everyone was improvising. There was no plan. This is especially true of “the blip.” I do wonder if there’s a way to incorporate that into Musk’s case. Anyway, Molo just referenced an exhibit he didn’t have handy, asked for an exhibit number, and then said he’d get it for the jury late.r I have to say, I know Musk’s team is smaller than OpenAI’s, but this might have been a moment to call in another lawyer to handle the close. Someone who could have prepped better, perhaps. Marc Toberoff, who’s theoretically a key figure on this team, hasn’t stood up to do a single thing. Maybe this could have been his moment, I don’t know!

Read on theverge.com →

AdventHealth advances whole-person care with OpenAI

By treating AI adoption as the outcome, AdventHealth eases clinician workload, improves workflows, and unlocks more time for patient care.

Results

80%

Reduction in time spent on administrative tasks

AdventHealth is deploying ChatGPT for Healthcare to reduce administrative burden and streamline clinical workflows across its system. By automating time-intensive documentation and support tasks, care teams are reclaiming hours each week and allowing clinicians to focus more directly on patients. The result is not just operational efficiency, but expanded clinical capacity, faster access to care, and a measurable improvement in the patient experience.

AdventHealth is a hospital system operating across nine states, serving millions of patients each year. Like many large health systems, it faces tight margins, growing demand and increasing administrative complexity.

Much of that pressure shows up in day-to-day workflows. Physician advisors reviewing cases for utilization management often spend about 10 minutes per case, not on a single task but on a sequence of steps: reading charts, identifying relevant details, checking criteria and drafting structured rationales. Across hundreds or thousands of cases, that time adds up quickly.

The burden extends beyond clinical roles. Teams in finance, HR, IT and other functions spend significant time drafting documents, summarizing information and preparing materials that are necessary but time-consuming. As a result, many operate in what leaders describe as “constant operations mode,” with limited capacity for higher-value work.

At the same time, interest in AI was already emerging inside the organization. Employees were experimenting with chatbots, even as formal policies restricted their use.

“We had folks who were eager to start, but there were a very large number of people who were on the sidelines,” says Rob Purinton, Chief AI Officer at AdventHealth. “They weren’t sure how to use AI effectively in their daily jobs.”

AdventHealth’s leadership concluded early that running isolated pilots would not lead to meaningful change. The central challenge was driving consistent, safe use across a large workforce.

“The hardest part of AI in healthcare is getting humans to use it safely, consistently, and at scale,” Purinton says. “We made a decision early on to treat adoption as the product.”

That decision shaped the rollout. Instead of positioning AI as automation, leaders framed it as a way to reduce administrative burden and return time to clinicians and staff.

“We don’t talk about AI as automation. We talk about time back,” Purinton says. “If we can take a 10-minute review and compress it meaningfully—while maintaining quality—that’s capacity we can give back to our clinicians.”

AdventHealth also treated adoption as a measurable operational metric. The organization tracks messages per user per business day, excluding weekends and holidays to create a consistent baseline. That metric is monitored and managed like any other KPI, with targets and trends reviewed regularly.

To scale usage, the system relied on domain-based peer groups rather than large, centralized training programs. Finance teams worked with finance teams and HR with HR, for example—sharing prompts, workflows and best practices relevant to their specific functions.

As the organization moved from experimentation to enterprise deployment, leadership prioritized tools that could meet healthcare requirements around privacy, governance and reliability.

“We chose OpenAI because we weren’t looking for a demo. We were looking for enterprise infrastructure,” Purinton says. “The reasoning capability, the structured outputs, and the governance controls gave us confidence that this wasn’t just productivity software. It was something we could responsibly scale across a health system.”

AdventHealth adopted ChatGPT Enterprise and later ChatGPT for Healthcare, which provided additional safeguards for regulated environments, including data protections and compliance support.

Speed of innovation and collaboration also influenced the decision.

“We really appreciate being closer to the edge of what’s possible,” Purinton says. “And we’ve found OpenAI to be highly collaborative as we think through pilots, deployments and what comes next.”

One of the earliest and most measurable use cases was utilization management.

Using ChatGPT for Healthcare, physician advisors can generate structured summaries of patient charts, surface relevant clinical details and draft initial rationales. The clinician remains responsible for final judgment, but the time spent assembling information is reduced.

The organization measures impact using system-level data, including timestamps in electronic health records, rather than self-reported estimates.

“We prefer measures that are baked right into the process,” Purinton says. “We can see exactly how many minutes have improved and whether that change is statistically significant.”

Beyond clinical workflows, similar patterns have emerged across departments:

- Drafting documents and plans starts with a first-pass output rather than a blank page

- Policies and communications are converted into structured, usable formats

- Notes and unstructured information are quickly summarized into action steps

These changes reduce cycle times, limit back-and-forth revisions and improve consistency in outputs.

AdventHealth evaluates AI impact across two primary dimensions: adoption and workflow performance.

On the adoption side, tracking daily usage has created accountability and visibility into how quickly AI is becoming part of routine work.

On the workflow side, pilots are evaluated using throughput metrics such as time per task, turnaround time and volume handled. In utilization management, the goal is to reduce review time while maintaining quality and consistency.

Across departments, teams report:

- Reduced time spent on repetitive documentation and review tasks

- Faster turnaround on internal workflows

- Fewer rework cycles due to more consistent first drafts

- Increased capacity without additional staffing

The organization often describes these gains as “time back,” but leadership ties that concept directly to measurable outcomes.

“If you take a 10-minute task and make it two, and that happens a thousand times a week, that’s real capacity,” Purinton says. “The question is how you reinvest that capacity.”

For AdventHealth, the value of AI is closely tied to its mission of delivering whole-person care. That requires time—time for clinicians to spend with patients and families, and time for staff to focus on higher-value work.

One example illustrates the impact at an individual level. A physician who previously spent evenings completing documentation, often referred to as “pajama time,” was able to finish work during regular hours after AI-supported changes to workflows.

“He was leaving work at work,” Purinton says. “He could go home and be present with his family.”

Stories like this reinforce the organization’s approach to AI as a tool for reducing administrative burden rather than replacing roles.

To date, most measurable gains have come from reducing time spent on existing tasks. AdventHealth views that as the starting point.

The organization is now focusing on expanding into areas such as patient access, clinical decision support and new care delivery models, while maintaining the same emphasis on governance, measurement and trust.

The core lesson, according to leadership, is that scaling AI depends less on the technology itself and more on how it is introduced and adopted.

“Adoption is not ‘go use the product.’ It’s ‘change leadership,’” Purinton says. “When you measure it, prove value and lead with trust, that’s when you get beyond pilots.”

Read on openai.com →

President Donald Trump has delayed signing an executive order that would allow the government to evaluate AI models before they’re released.

Trump claimed he is not happy with the language of the order: “I didn’t like certain aspects of it,” he told the White House press pool. “We’re leading China, we’re leading everybody, and I don’t want to do anything that’s going to get in the way of that leading.”

The unofficial reason: Not enough tech CEOs could make it to Washington, D.C. on short notice, according to several reports. And what’s an executive order signing without a photo op?

The anticipated executive order would have tasked the Office of the National Cyber Director and other agencies with developing a process to evaluate AI models for security before their release. This is partly in response to concerns from the release of Anthropic’s Mythos and OpenAI’s GPT-5.5 Cyber — both of which can quickly find and exploit security vulnerabilities.

One of the key sticking points in the EO’s language, per CNN, is a proposed requirement for AI companies to share advanced models with the government between 14 and 90 days ahead of launch.

Trump said he was concerned that the EO’s language today “could have been a blocker.”

Read on techcrunch.com →

What will it take to launch the first must-have AI consumer product? Maybe $700 million.

At least according to Hark, an AI lab building models and hardware for an AI personal assistant, which said on Thursday that it had raised that much in a Series A round that values it at $6 billion post-money.

The mega-round was led by Parkway Venture Capital and included Nvidia, Align Ventures, AMD Ventures, ARK Invest, Brookfield, Greycroft, Intel Capital, Prime Movers Lab, Qualcomm Ventures, Salesforce Ventures, and Tamarack Global. (Phew!)

Perhaps what’s most notable about the fundraise is how little Hark has revealed about what it is building. Founder and CEO, Brett Adcock, also the entrepreneur behind robotics company Figure.AI and electric aircraft builder Archer, launched Hark in late 2025 with $100 million of his own money to develop an agentic AI system that serves as a universal interface with the digital world.

Hark expects to release its first multimodal models this summer, which it says will power a personal AI platform that works with existing products and services. The company expects to follow that with hardware devices built specifically for those systems.

The fresh cash will be spent on recruiting top talent for hardware, product design, and AI research, and on securing compute and components. The company currently has 70 employees and runs a data center with Nvidia B200 GPUs.

Abidur Chowdhury (pictured above in a promo video), a former Apple product executive, is Hark’s director of design. He declined to reveal new details of what he’s working on when TechCrunch peppered him with questions this week but said investors were impressed by a series of demos from his team.

“I haven’t seen anything that feels like something that will really help like the normal person,” Chowdhury said, speaking of the AI products on the market. “People are really building things to help people make software, and it’s working, and it’s really impactful, but we haven’t really seen that for the normal person yet.”

He noted that while Anthropic is prioritizing coding tools and OpenAI is moving in the same direction ahead of its IPO, few companies are focused solely on building interfaces and native hardware the way Hark is. “With this focus, with this great team that we have, and this round that we’ve raised, I think we can make something really special in this space,” Chowdhury said.

Still, there are more questions than answers. One challenge will be providing the context of a customer’s life to an AI assistant without making the people around the user uncomfortable or violating their privacy. Wearables like Meta’s existing glasses or the forthcoming Android spectacles don’t seem to have solved this problem. When asked how he might square this particular circle, Chowdhury only smiled.

“Sounds like that would make a great product.”

Read on techcrunch.com →