To most consumers, Uber is a ride-hailing and food delivery app. But inside the company, it increasingly sees itself as something else: a living, evolving codebase.
That framing comes directly from CEO Dara Khosrowshahi, who recently described Uber’s engineers as “literally the builders of the company.” And in a twist that captures just how deeply artificial intelligence is embedding itself into corporate life, some of those builders have gone a step further, and they created an AI version of their own CEO.
Yes, Uber engineers built a chatbot modelled after Khosrowshahi. And teams are using it to rehearse presentations before stepping into meetings with the real ones.
Khosrowshahi revealed the details during an interview on The Diary of a CEO podcast, describing how employees present slide decks to “Dara AI” as a dry run before facing executive leadership.
“You can imagine,” he said, “by the time something comes to me, there’s been prep and a meeting, and the slide deck has been beautifully honed. So they have Dara AI to tune their prep.”
What sounds like a Silicon Valley inside joke is actually a telling signal of something larger: AI at Uber is no longer just a backend optimization engine. It’s becoming part of the company’s cultural operating system.
According to Khosrowshahi, about 90% of Uber’s software engineers now use AI in their daily workflows. Roughly 30% qualify as what he calls “power users” — engineers who are not just accelerating tasks, but fundamentally rethinking Uber’s architecture through AI-assisted development.
“They are manufacturing the bricks that go into the system,” he said, “and they’re architects thinking about what the system should look like.” That shift matters.
Across corporate America, AI integration often follows a top-down script: executives approve tools, IT teams deploy them, and employees experiment cautiously. At Uber, the Dara chatbot appears to represent the opposite of organic experimentation bubbling up from engineers themselves.
Instead of waiting for corporate mandates, teams built a bespoke tool to solve a specific workplace friction point: executive scrutiny.
Presenting to a CEO is rarely casual. There are performance stakes, strategic implications, and reputational risks. By simulating Khosrowshahi’s questioning style and decision logic, engineers created a low-stakes rehearsal room powered by generative AI.
It’s unclear exactly how the chatbot was trained, but such systems can draw on public interviews, internal communications, leadership memos, and historical Q&A patterns to approximate tone and priorities.
The result is something closer to a leadership simulator than a novelty bot.
And it raises interesting questions. If teams can anticipate executive feedback with algorithmic precision, does that improve clarity or risk narrowing creative debate? Does AI-assisted rehearsal strengthen ideas or standardize them? Either way, it signals that AI is reshaping not only how code gets written, but how influence gets navigated.
Uber has long been an AI-heavy company. Its logistics engine relies on machine learning for route optimization, dynamic pricing, fraud detection, and demand forecasting. Autonomous vehicle research and marketplace balancing have pushed the company deeper into advanced modelling for years, but this moment is different.
The Dara chatbot reflects AI moving from operational infrastructure to interpersonal infrastructure, the softer, more human dimensions of corporate life: communication, persuasion, hierarchy.
In many companies, AI still lives in dashboards and developer consoles. At Uber, it is beginning to mediate leadership dynamics.
And Khosrowshahi’s reaction is telling. Rather than resisting or downplaying the project, he framed it as part of a broader productivity revolution. “It really is changing their productivity in a way that I’ve never, ever seen before,” he said.
Read also: Uber Pilots In-App Video Recording for Drivers Across India
As generative AI becomes more sophisticated, the tools to simulate communication styles, decision-making logic, and leadership personas are increasingly accessible. Uber may simply be early in what becomes a broader workplace trend.
Imagine AI simulations for board presentations. Investor pitch rehearsals. Regulatory hearings. Sales negotiations.
What Uber’s engineers built for internal use could evolve into a new category of enterprise tooling: executive simulation platforms.
There’s also a subtle power dynamic at play. Turning a CEO into a chatbot requires a level of cultural confidence and psychological safety. Not every company would embrace such experimentation. That Uber’s leadership appears comfortable with it suggests a strategic embrace of AI as both a tool and a transformation driver.
In a tech industry racing to prove meaningful AI integration beyond marketing slogans, Uber’s internal experiment offers something rare: a bottom-up case study in real-world adoption.
The company that once disrupted transportation may now be quietly prototyping something else, the AI-mediated corporation and if engineers are already rehearsing with a digital CEO, the next frontier may not be replacing jobs with AI but redefining how leadership itself gets experienced.
