A high profile OpenAI departure points straight at the Pentagon deal and shows the storm is not blowing over

Good morning. It’s amazing how many of the top tech business stories right now feature Oracle playing either a central role, or some level of supporting role. There’s the American TikTok joint venture, the OpenAI data center project (part of which is reportedly being scaled back), the $111 billion Paramount Warner Bros acquisition (with a fair chunk of the price guaranteed by Larry Ellison’s Oracle equity), and the tech industry layoffs (reports emerged last week that Oracle plans to cut thousands of jobs), and the Big Tech bond binge (see news item below).

For a behind-the-scenes enterprise software company, Oracle sure knows how to get in the news.

This week, the spotlight will shine directly on Oracle when it reports its quarterly earnings after the bell on Tuesday. In addition to all the updates on Oracle’s actual business, Wall Street has lots of questions about the various projects that Oracle and its cofounder are at the center of. It should be an interesting conference call.

Today’s tech news below.

Alexei Oreskovic
@lexnfx
alexei.oreskovic@fortune.com

Want to send thoughts or suggestions to Fortune Tech? Drop a line here.

OpenAI robotics leader resigns over Pentagon deal concerns

More aftershocks of OpenAI’s Pentagon deal hit the tech industry over the weekend as Caitlin Kalinowski, who had been leading hardware and robotics operations teams at OpenAI since November 2024, announced she has left the company due to concerns about the deal.

“I resigned from OpenAI,” she posted on X and LinkedIn on Saturday. “I care deeply about the Robotics team and the work we built together. This wasn’t an easy call. AI has an important role in national security. But surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got."

Her departure comes amid an escalating dispute over how far AI companies should go in supporting U.S. military uses of the technology. Talks between Anthropic and the Pentagon collapsed after the company pushed for strict limits on domestic surveillance and autonomous weapons. OpenAI’s deal with the Department of Defense has drawn criticism from some employees and observers who argued that OpenAI stepped in after Anthropic refused the terms.

Before joining OpenAI, Kalinowski had spent several years leading Meta’s AR glasses effort and working on its Oculus headsets, and worked at Apple before that. “This was about principle, not people,” Kalinowski wrote. "I have deep respect for Sam and the team, and I’m proud of what we built together.​​​​​​​​​​​​​​​​”—Sharon Goldman

Big Tech’s pivot to debt

Much of the capital expenditures for the AI data center buildout is being paid for with operating cash flows. But the sheer magnitude of the spending has prompted many of the internet companies to do something they haven’t done much of in the past: borrow money.

In 2025, Alphabet, Amazon, Oracle, Meta and Microsoft issued about $121 billion in new debt via bonds, compared to $40 billion in 2020. And the pace is not expected to slow down anytime soon: Wall Street estimates show the AI-related bond supply could be in the range of $100 billion to $300 billion this year. Over the next three to five years, total data center investment could run $1.5 trillion to $3 trillion, according to some analyses.

The trend is already changing the stakes for businesses that have traditionally had no need to borrow, introducing a new layer of stakeholders, obligations, and risks that are transforming how internet companies operate and how they are valued by investors. Bond investors, unlike equity investors, don’t seek out unlimited upside, they focus on being compensated fairly for taking on risks, including those related to overinvestment that leads to a glut in supply.  Read Fortune’s deep dive on the AI bond binge and why it matters.—Amanda Gerut

The AI Jobs Gap

A new study from Anthropic researchers has put hard numbers on how close AI could be to displacing professional work.

Drawing on real Claude usage data, the paper introduces a metric called “observed exposure,” measuring what AI is theoretically capable of against what it’s actually being used for in professional settings. According to the study, while large language models could theoretically handle 94% of tasks in computer and math roles, they currently cover only around a third in observed use.

The highest-risk group is more often female, highly educated, and well-compensated, with roles like lawyers, financial analysts, and software developers bearing far more exposure than many manual or in‑person roles such as cooks, bartenders, or mechanics.

The study echoes a growing chorus of warnings from top tech leaders. Microsoft AI CEO Mustafa Suleyman recently told the Financial Times that most tasks involving sitting at a computer—law, accounting, marketing, project management—could be largely automatable on roughly a one‑to‑two‑year time frame, while Anthropic’s own CEO Dario Amodei has separately warned that AI could wipe out up to about half of all entry‑level white‑collar jobs.

Legal constraints and the ongoing need for human oversight are keeping adoption low for now, but Anthropic researchers say those are more likely to be temporary rather than structural barriers.—Beatrice Nolan

More tech

—Samsung wants to do more AI deals. Open to ‘strategic partnerships’

—China’s Tencent in the Paramount deal? A passive financial investor.

—About that SpaceX IPO… How likely is it to justify a $1.5 trillion market cap

—AI ‘brain fry’. It’s a thing, per a new HBR report.

—‘Grandfather of the Internet’ dies at 91. RIP David J. Farberg

This is the web version of Fortune Tech, a daily newsletter breaking down the biggest players and stories shaping the future. Sign up to get it delivered free to your inbox.