Mira Murati breaks silence with AI governance warning and new interaction models at Thinking Machines

Mira Murati steps back into the spotlight, carefully

Mira Murati isn't a natural creature of the conference stage. As the CTO of OpenAI, she was present but rarely the public face of the company. As CEO of her own company, Thinking Machines Lab, she has been even harder to find. So when she sat down with Bloomberg in San Francisco on Thursday -- her first major public appearance in roughly 18 months -- it was worth paying attention, even if she was careful not to say too much. The timing makes sense. Thinking Machines has spent the better part of a year and a half operating largely in the background: raising capital, hiring researchers, and shipping one product, Tinker, an API for fine-tuning open-source AI models. In the meantime, the companies competing for the same talent, customers, and headlines have only grown more omnipresent. OpenAI, where Murati spent six years as CTO, is constantly in the news cycle. Anthropic's momentum is all that anyone can talk about right now. And xAI, Elon Musk's AI venture, has been folded into SpaceX ahead of what is expected to be its massive public offering, generating its own gravitational pull on attention and investment. In that environment, staying heads down has diminishing returns; at some point, you have to make some noise just to remind the market you exist. Murati used the Bloomberg appearance to do exactly that and not much more. She previewed what Thinking Machines is calling "interaction models," which she described as a fundamentally different kind of AI interface. Rather than the turn-based, prompt-and-response dynamic that defines most AI products today, she told interviewer Emily Chang, the company's models are designed to process continuous streams of audio, text, and video in 200-millisecond intervals. The idea is that they can pick up on the texture of human communication -- the interruptions, the mid-thought corrections, even pauses to think -- in something closer to real time. It fits the lab's core thesis that the path to powerful AI runs through closer human collaboration, not around it. But Murati was careful to frame it as a first step, not a finished product, and she declined to put a specific release date on anything. She also answered questions about the episode that first put her more squarely in the public eye: the chaotic week in November 2023 when OpenAI's board fired Sam Altman and she became interim CEO. Inside OpenAI it came to be called "the blip." Murati said she felt clear about her decisions in each moment -- that protecting the mission and the team was the through-line that made the choices feel obvious even as the situation appeared to be falling apart from the outside. She said the company would have "imploded" if not for her involvement through that strange five-day stretch and its immediate aftermath. But she acknowledged that clarity of intent is not the same thing as clarity about consequences. In retrospect, she said, she would have pushed harder for more information, a better transition plan, and more transparency. What she did not say, at least not directly, is whether she thinks things turned out well. Asked whether she still trusts her former boss, she sidestepped the question, steering the conversation toward a larger concern that she returned to several times: the concentration of consequential decisions in too few hands -- not just at OpenAI but across the industry. Her worry, she said, is less about the character of any individual leader (though she acknowledged that matters) and more about the absence of structural checks. Good people make bad calls. Well-intentioned organizations drift. Too much attention has been paid to virtue and too little to governance, she suggested. Chang also politely pressed her on the departures of several high-profile researchers from Thinking Machines in recent months , a subject Murati has largely avoided in public and that she downplayed on Thursday. First, she said, building a frontier AI lab from scratch compresses years of normal organizational volatility into months. She also acknowledged that compensation -- the nine-figure packages that have become standard currency in the war for AI talent -- captures people's imaginations, but she suggested it isn't usually the whole story. To some audience laughter, she said of her own competitive instincts, "When I wake up in the morning, I am not thinking about how to kill the competitor." Naturally, Chang asked about what comes next for AI broadly, including for the humans who AI companies once said would be empowered by AI but who've more recently grown scared by talk of mass job displacement, not to mention a future where AI is used to create chemical weapons. Murati, who was born in Albania and speaks with a slight Eastern European accent, was measured in her response. She pushed back on the framing of inevitable dystopia or inevitable utopia, arguing that neither outcome is predetermined and that the period we're in right now is the one that will determine which way things go. Still, she said -- and not for the first time during the interview -- that if humans take their hands off the wheel too soon, the future will look very different, and not better.

Mira Murati resurfaces with AI governance warning, new product

Former OpenAI CTO Mira Murati made her first major public appearance in 18 months, previewing Thinking Machines Lab's "interaction models" and arguing that the AI industry lacks structural governance checks. She also addressed researcher departures and reflected on the 2023 Altman firing. For someone who helped ship ChatGPT, DALL-E, and Codex, Mira Murati has been remarkably quiet. On Thursday, she broke the silence. Sitting down with Bloomberg's Emily Chang in San Francisco, the CEO of Thinking Machines Lab gave her first major media appearance in roughly 18 months, a carefully managed re-entry into a conversation that has moved at breakneck speed without her. The timing was not accidental. Thinking Machines has spent that year and a half raising $2 billion, securing a gigawatt of Nvidia Vera Rubin compute, shipping one product, and losing a troubling number of the researchers it hired to build the next one. The AI landscape Murati left behind when she departed OpenAI in September 2024 looks nothing like the one she re-entered on Thursday. The product: interaction models Murati used the appearance to preview what Thinking Machines is calling "interaction models," a fundamentally different kind of AI interface. Rather than the prompt-and-response format that defines most AI products, the company's models are designed to process continuous streams of audio, text, and video in 200-millisecond intervals. The pitch is that these models can pick up on the texture of human communication: interruptions, mid-thought corrections, pauses. The technical term is "full duplex," and the company claims its TML-Interaction-Small model responds in 0.40 seconds, roughly the speed of natural conversation. It fits Thinking Machines' founding thesis that powerful AI requires closer human collaboration, not less of it. Murati was careful to frame this as a first step. She declined to put a release date on anything, and she positioned the work alongside Tinker, the company's API for fine-tuning open-source models, which launched in October 2025 and remains its only shipping product. The departures Chang pressed Murati on what has quietly become the company's most visible problem: a string of high-profile departures. Co-founder and CTO Barret Zoph, co-founder Luke Metz, and founding team member Sam Schoenholz all returned to OpenAI in January. Five founding members have gone to Meta, reportedly lured by compensation packages that reach into nine figures. Murati downplayed the exits. Building a frontier AI lab from scratch compresses years of normal organisational volatility into months, she said. She acknowledged that the nine-figure packages now standard in the AI talent war capture imaginations, but suggested compensation is rarely the whole story. "When I wake up in the morning, I am not thinking about how to kill the competitor," she said, drawing laughter from the audience. The line was disarming, but the competitive reality is stark. OpenAI is everywhere. Anthropic has raised $30 billion and reportedly attracted investor offers at an $800 billion valuation. Elon Musk's xAI has been folded into SpaceX ahead of a record IPO. In that environment, staying quiet has costs. The Altman firing, revisited Chang asked about the episode that first made Murati a public figure: the chaotic five days in November 2023 when OpenAI's board fired Sam Altman and Murati became interim CEO. Inside OpenAI, the incident came to be called "the blip." Murati said she felt clear about her decisions in each moment, that protecting the mission and the team was the thread that made the choices feel obvious even as the situation looked like it was falling apart from outside. She said the company would have "imploded" without her involvement through that stretch. But she acknowledged that clarity of intent is not the same as clarity about consequences, and said she would have pushed harder for more information, a better transition plan, and more transparency. Asked whether she still trusts Altman, she sidestepped. What she offered instead was more interesting: a broader argument about the concentration of consequential decisions in too few hands, not just at OpenAI but across the industry. Her concern, she said, is less about the character of any individual leader and more about the absence of structural checks. Good people make bad calls. Well-intentioned organisations drift. The harder question On the future of AI broadly, Murati pushed back on both the dystopian and utopian framings. Neither outcome is predetermined, she argued. The period we are in right now is the one that will determine which way things go. But she returned, more than once, to a theme that connects her governance critique to her product philosophy: if humans take their hands off the wheel too soon, the future will look very different, and not better. It is a position that sits comfortably with her company's thesis about human-AI collaboration. Whether it can survive contact with a market that rewards speed, scale, and tens of billions in capital over caution is the question Murati did not answer on Thursday. She does not need to answer it yet. But with one product shipping, a team that keeps shrinking at the top, and competitors that grow louder by the week, the window for quiet conviction is closing.

Thinking Machines' Mira Murati bats for responsible innovation, human connect

In an interview with Bloomberg, Murati said the decision to start the company stemmed from a desire to build a frontier AI lab focused on human-AI collaboration. Thinking Machines is building AI interaction models that make human-AI conversations rich and nuanced in real time instead of "turn-based". Artificial intelligence (AI) is entering a new phase, where machines are not just tools but also collaborators, says Mira Murati, cofounder and CEO of Thinking Machines Lab. In an interview with Bloomberg, Murati said the decision to start the company stemmed from a desire to build a frontier AI lab focused on human-AI collaboration. Thinking Machines is building AI interaction models that make human-AI conversations rich and nuanced in real time instead of "turn-based". Murati said that while it is difficult to predict the future, "the most important thing would be to feel a sense of agency and possibility of the future." Towards human-like interaction Elaborating on Thinking Machines' new "interaction models", Murati described a future where speaking with AI "feels more like speaking with another person and less like sending messages back and forth." She said the company is developing systems that can process incoming information and generate responses simultaneously, enabling more natural conversations. Speaking on the cutthroat competition in the ever-evolving AI sphere, Murati said, "Competition is good as it leads to better products and technologies for people." While the competition is fierce, it is also important to balance quick action with durable long-term progress, she added. Thinking Machines Lab's interaction model is currently in research preview and not publicly available. The company says it will release a limited preview in the coming months, with wider access planned later this year. The company says its approach is meant to keep humans involved throughout the interaction, rather than pushing them out of the process. OpenAI recollections Murati was the chief technology officer of OpenAI before founding Thinking Machines Lab. Asked about the OpenAI board change, she said that while the situation looked "chaotic" from the outside, she stood by the feedback she had delivered to the board. "In retrospect, I think I would have paused more to understand the transition plan," she added. Murati was appointed interim CEO of OpenAI after its board temporarily forced out cofounder Sam Altman in ⁠2023. She recently testified in a federal court that Altman was "creating chaos" and, at times, was deceptive with her and others. Murati testified that Altman pitted executives against one another and undermined her role as technology chief. In the interview, Murati called for improved checks and balances on AI governance. "It's tempting to push forward rapidly, but we need to build trust," she told Bloomberg. "That means clear communication, external audits, and collaboration with policymakers."