Emil Michael, who serves as a senior technology official at the Department of Defense, is back in the spotlight over the government’s ongoing battle with Anthropic, and a newly released podcast interview offers one of the most detailed looks yet into his thinking on that dispute — as well as an unguarded settling of old scores from his Uber days.
The interview, released Monday and conducted last month by Joubin Mirzadegan, a partner at Kleiner Perkins who leads the venture firm’s portfolio operating team, covered a range of topics including policy and personal history — and was recorded before the DoD’s feud with Anthropic had fully come to a head. But it is Michael’s remarks about his departure from Uber — and his barely concealed bitterness about it — that grabbed our attention first.
When Mirzadegan asked him point-blank whether he had been shown the door alongside Travis Kalanick, Michael answered with a single word: “Effectively.”
Michael resigned eight days before Kalanick did in June of 2017, as part of the fallout from a workplace investigation triggered by allegations of sexual harassment and gender discrimination at the company. He was not named in those allegations, but the inquiry — led by former U.S. Attorney General Eric Holder — concluded he should be removed. Kalanick followed, pushed out in what The New York Times described as a shareholder revolt by some of the company’s most prominent investors, including Benchmark.
When Mirzadegan asked whether he was still “salty” about it, Michael didn’t equivocate. “I’ll never forget that, nor forgive,” he said.
The ouster grates on both Michael and Kalanick not only because of the personal damage to their reputations but because they believed — and still believe — that autonomous driving was Uber’s future, and that the investors who forced them out killed it.
During the interview, Michael argued the decision was driven by a desire to protect near-term returns rather than build something lasting.
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
“They wanted to preserve their embedded gains, rather than try to make this a trillion-dollar company,” he said.
Kalanick has been equally pointed. At the Abundance Summit in Los Angeles last year, he said the program was second only to Waymo at the time of its cancellation and closing the gap. “You could say, ‘Wish we had an autonomous ride-sharing product right now. That would be great,’” he told the audience.
Uber sold its self-driving unit to Aurora in what was widely perceived as a fire sale in 2020, three years after both men were gone. The decision looked defensible at the time; autonomous driving was burning cash, and the tech felt very distant. Now Waymo’s robotaxis are operating in 10 U.S. cities and expanding into new markets. Whether Uber ever had the staying power to get there is an open question, but it’s clearly one that still haunts both men.
For his part, Kalanick never really stopped building. This month he took the wraps off Atoms, a robotics company he has been developing in stealth since around the time he left Uber eight years ago. He also revealed he is the largest investor in Pronto, an autonomous vehicle startup focused on industrial and mining sites founded by his former Uber colleague Anthony Levandowski, and said he is on the verge of acquiring it outright.
Meanwhile, Michael has found a new battlefront. The interview was recorded just before the DoD’s negotiations with Anthropic publicly collapsed, and his account of that standoff is worth a listen. He describes Anthropic as one of only a handful of approved large language model vendors for the department, approved in part through its partnerships with Palantir. As Michael frames it, the DoD is hardly a free-for-all. It operates under such a dense web of laws, regulations, and internal policies that “we almost choke on them,” he tells Mirzadegan. Anthropic, he argues, wants to add its own layer on top of all of that.
“What I can’t do is have any one company impose their own policy preferences on top of the laws and on top of my internal policies,” he said, using an analogy to make his point. “If you buy the Microsoft Office Suite, they don’t tell you what you could write in a Word document, or what email you can send.”
Michael then went further, invoking a finding Anthropic itself had published last month ahead of his conversation with Mirzadegan. Chinese technology companies, he argued, had been hitting Anthropic’s models repeatedly in a technique called distillation — essentially reverse-engineering the model’s behavior closely enough to replicate its capabilities.
Through China’s civil-military fusion laws, he said, that would give the People’s Liberation Army access to something functionally equivalent to Anthropic’s full, unrestricted model. Meanwhile, the DoD would be working with a version hemmed in by Anthropic’s own guidelines. “I’d be one-armed, tied behind my back against an Anthropic model that’s fully capable — by an adversary,” Michael said. “It’s totally Orwellian.”
Michael added a bit later in the interview, before moving on to the next topic: “If you’re an American champion — and I believe they are, they’re one of the most important companies in the country — don’t you want to help your Department of War succeed with the best tools available?”
As industry watchers are well aware, the dispute has since moved from negotiating table to courtroom.
Defense Secretary Pete Hegseth deemed Anthropic a “supply-chain risk” in late February, and the government escalated further last week, filing a 40-page brief in U.S. District Court for the Northern District of California. The brief argued that giving Anthropic access to the DoD’s war-fighting infrastructure would introduce “unacceptable risk” into its supply chains in part because the company could theoretically disable or alter its own technology to suit its interests rather than the country’s in a time of war.
Anthropic fired back on Friday, submitting sworn declarations, alongside a brief, arguing the government’s case rests on technical misunderstandings and claims that were never raised during months of prior negotiations. One of those declarations, filed by Anthropic’s head of public sector Thiyagu Ramasamy, directly challenged the government’s claim that Anthropic could interfere with military operations by disabling or altering how its technology behaves — something Ramasamy says is not technically possible.
A hearing is scheduled for Tuesday in San Francisco.
