By Titik Puspa 
Emil Michael, a senior technology official within the Department of Defense, has re-emerged into the public consciousness, not just for his current role in the government’s contentious standoff with AI firm Anthropic, but also for a recent, remarkably candid podcast interview. This conversation, released this past Monday, offers an unprecedented deep dive into Michael’s strategic thinking on the Anthropic dispute and, perhaps more unexpectedly, provides an unguarded airing of grievances stemming from his tumultuous departure from Uber. The interview, recorded last month by Joubin Mirzadegan, a partner at venture capital firm Kleiner Perkins and head of its portfolio operating team, touched upon a wide array of subjects, encompassing both policy and personal history. Crucially, it predates the full escalation of the Department of Defense’s (DoD) conflict with Anthropic, lending a unique perspective to Michael’s commentary. However, it is his reflections on his exit from Uber, marked by a palpable and barely concealed bitterness, that first captured significant attention.
When directly questioned by Mirzadegan about whether he was effectively pushed out alongside former Uber CEO Travis Kalanick, Michael’s response was stark and definitive: "Effectively." This admission underscores the deep-seated resentment that lingers from an event that fundamentally reshaped the leadership of one of the world’s most disruptive technology companies. Michael’s resignation occurred a mere eight days prior to Kalanick’s own departure, a consequence of a rigorous workplace investigation. This investigation was instigated by serious allegations of sexual harassment and gender discrimination that plagued Uber’s corporate culture. While Michael himself was not directly implicated in these allegations, the inquiry, spearheaded by former U.S. Attorney General Eric Holder, ultimately recommended his removal. Kalanick followed suit, forced out in a move widely characterized by The New York Times as a shareholder revolt, orchestrated by some of Uber’s most influential investors, including the prominent venture capital firm Benchmark.
Mirzadegan’s probing question about whether Michael harbored lingering resentment, asking if he was still "salty" about the situation, was met with an unequivocal response. "I’ll never forget that, nor forgive," Michael stated, leaving no room for doubt about the enduring impact of his ouster. This painful expulsion from Uber resonates deeply with both Michael and Kalanick, not solely due to the damage inflicted upon their professional reputations, but more profoundly because of their shared conviction that autonomous driving represented Uber’s true future. They firmly believed, and continue to believe, that the investors who engineered their departure effectively extinguished this crucial, forward-looking initiative.
During the podcast interview, Michael articulated his perspective that the decision to divest from autonomous driving was driven by a shortsighted desire to safeguard immediate financial returns, rather than to cultivate a lasting, transformative enterprise. "They wanted to preserve their embedded gains, rather than try to make this a trillion dollar company," he asserted, painting a picture of a company prioritizing short-term financial gratification over long-term strategic vision.
Travis Kalanick has echoed this sentiment with equal vehemence. At the Abundance Summit held in Los Angeles the previous year, Kalanick highlighted the advanced state of Uber’s autonomous driving program at the time of its cancellation, asserting it was second only to Waymo and actively closing the competitive gap. He remarked to the audience, "You could say, ‘Wish we had an autonomous ride-sharing product right now. That would be great.’" This statement reflects a profound sense of missed opportunity and strategic miscalculation on Uber’s part.
The eventual fate of Uber’s self-driving unit saw it sold to Aurora in 2020, three years after Michael and Kalanick had departed. This transaction was widely perceived as a "fire sale," a desperate attempt to offload a struggling asset. At the time, the decision appeared justifiable. The autonomous driving sector was a significant drain on capital, and the technological hurdles seemed immense and distant. However, the landscape has dramatically shifted. Waymo’s robotaxis are now operational in ten U.S. cities and are actively expanding into new markets, demonstrating the viability and accelerating progress of the technology. Whether Uber possessed the strategic foresight and financial resilience to navigate this complex journey remains an open question, but it is a question that clearly continues to haunt both Michael and Kalanick.
Kalanick, for his part, has not abandoned the pursuit of innovation. In a significant development this month, he unveiled Atoms, a robotics company he has been developing in stealth mode for approximately eight years, since his departure from Uber. Furthermore, he revealed his position as the largest investor in Pronto, an autonomous vehicle startup focused on industrial and mining applications. Pronto was founded by his former Uber colleague, Anthony Levandowski, and Kalanick indicated he is on the verge of acquiring the company outright, signaling a continued commitment to the autonomous vehicle space.
Meanwhile, Emil Michael has found himself embroiled in a new and complex battlefront: the Department of Defense’s intricate relationship with artificial intelligence, specifically with the AI firm Anthropic. The interview was conducted shortly before the public collapse of negotiations between the DoD and Anthropic, making his account of that standoff particularly compelling. Michael described Anthropic as one of a select group of approved large language model (LLM) vendors for the department, a status partially achieved through its collaborations with Palantir. He emphasized that the DoD is far from a free-for-all environment. Instead, it operates under an extraordinarily dense framework of laws, regulations, and internal policies that, in his words, "we almost choke on them." Michael’s central contention is that Anthropic is attempting to superimpose its own policy preferences atop this already complex regulatory structure.
"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," Michael stated, employing a vivid analogy to illustrate 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." This analogy effectively highlights his concern that Anthropic’s restrictions could unduly limit the DoD’s operational flexibility and strategic autonomy.
Michael further amplified his concerns by referencing a recent finding published by Anthropic itself, shortly before his interview. He pointed to evidence suggesting that Chinese technology companies had been repeatedly targeting Anthropic’s models through a technique known as "distillation." This process, Michael explained, involves reverse-engineering the model’s behavior to a degree that allows for the replication of its capabilities. Under China’s civil-military fusion laws, Michael argued, this would grant the People’s Liberation Army access to a capability functionally equivalent to Anthropic’s full, unrestricted model. In stark contrast, the DoD would be constrained to using a version subject to Anthropic’s own, more restrictive guidelines. "I’d be one-armed, tied behind my back against an Anthropic model that’s fully capable – by an adversary," Michael declared, characterizing the situation as "totally Orwellian." This framing suggests a profound national security risk, where an adversary could possess unfettered access to advanced AI capabilities while the U.S. military is deliberately hobbled.
Adding a layer of urgency to his plea, Michael concluded this segment of the interview with a direct appeal. "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?" This question underscores his belief that American technology companies should prioritize national security interests and collaborate closely with the DoD to ensure the nation’s technological superiority.
The dispute between the DoD and Anthropic has since transitioned from the negotiating table to the courtroom, a development that industry observers have closely followed. In late February, Defense Secretary Pete Hegseth officially designated Anthropic as a "supply-chain risk." The government further escalated the matter last week by filing a comprehensive 40-page brief in the U.S. District Court for the Northern District of California. This legal filing argued that granting Anthropic access to the DoD’s war-fighting infrastructure would introduce "unacceptable risk" into its supply chains. The brief specifically raised concerns that Anthropic could, in a time of conflict, theoretically disable or alter its own technology to serve its interests rather than those of the United States, thereby compromising national security.
Anthropic responded forcefully on Friday, submitting sworn declarations alongside its own brief. The company contested the government’s case, asserting that it rests on technical misunderstandings and claims that were never raised during months of prior negotiations. One of these declarations, filed by Thiyagu Ramasamy, Anthropic’s head of public sector, directly challenged the government’s assertion that Anthropic could interfere with military operations by disabling or altering its technology’s behavior. Ramasamy stated unequivocally that such an action is not technically feasible, directly refuting a core tenet of the DoD’s legal argument.
A pivotal hearing in this complex legal battle is scheduled for Tuesday in San Francisco, where the arguments of both the Department of Defense and Anthropic will be further scrutinized by the court. This legal showdown promises to have significant implications for the future of AI procurement and deployment within the U.S. military, as well as for the broader landscape of AI regulation and national security in the digital age. The outcome will not only shape the relationship between government and cutting-edge technology firms but will also shed light on the delicate balance between fostering innovation and mitigating perceived risks in a rapidly evolving technological environment.