# Sam Altman [Frontier Minds profile](https://frontierminds.ai/people/sam-altman/) Last verified: 2026-09-19 Positions and narratives are editorial summaries. Quotes are attributed quotations. Source dates and profile verification dates are distinct. Missing positions mean not recorded, not agreement or neutrality. CEO · OpenAI San Francisco, USA. Born 1985 Co-founder and CEO of OpenAI, the company behind ChatGPT, and president of Y Combinator from 2014 to 2019; the executive most identified with the drive to build and deploy artificial general intelligence. ## Overview ### Early life and education Samuel Harris Altman was born on 22 April 1985 in Chicago, the eldest of four children of a dermatologist, Connie Gibstine, and a real estate broker, Jerry Altman. The family moved to Jerry's hometown of Clayton, Missouri, a suburb of St. Louis, in 1989. Sam attended John Burroughs School, a private school in nearby Ladue. He came out to his parents at 16, and when a Christian group boycotted a school assembly on sexuality, he addressed the whole school, announced that he was gay, and asked whether Burroughs wanted to be a repressive place or one open to different ideas. He enrolled at Stanford to study computer science and left after two years, in 2005, when he and two classmates dropped out to work full time on a mobile app. Peter Thiel, who became his mentor and main backer, later described him as "an optimist yet a survivalist, with a sense that things can always go deeply wrong." Altman told The New Yorker in 2016 that he kept "guns, gold, potassium iodide, antibiotics, batteries, water, gas masks from the Israeli Defense Force, and a big patch of land in Big Sur I can fly to." ### Loopt and Y Combinator The app was Loopt, which told your friends where you were. It was one of the eight companies in Y Combinator's first batch in the summer of 2005, when Paul Graham gave each founder six thousand dollars, advice and home-cooked chicken fricassee. Graham wrote in 2008 that "you could parachute him into an island full of cannibals and come back in 5 years and he'd be the king." Loopt never found a mass audience. Green Dot, a prepaid debit card company, bought it in 2012 for about 43 million dollars, a sale that TIME later reported netted Altman about 5 million dollars. He put part of that into a small venture fund, Hydrazine Capital, with his brother Jack, and found that he disliked venture capital. He had been a part-time YC partner since 2011. On 21 February 2014 Graham announced that Altman would become president "starting next batch," writing that "YC needs to grow, and I'm not the best person to grow it." Under Altman the accelerator added a fund and program for larger companies, expanded into China, and pushed into energy, biotech and other hard-technology sectors; Helion, the fusion company he later backed with 375 million dollars of his own money, was among the first. In November 2014 he spent eight days as interim chief executive of Reddit after Yishan Wong resigned, writing on his blog that the stint had been "sort of fun" and that he was happy it was ending. He left YC in March 2019, the same week OpenAI announced its restructuring, and formally became OpenAI's CEO. In 2024 Graham denied a persistent rumor that YC had fired him, saying that once OpenAI created a for-profit arm with Altman as CEO, YC's Jessica Livingston told him he had to choose one job. ### Founding and leading OpenAI In the summer of 2015 Altman tracked down Ilya Sutskever, then a star researcher at Google Brain, and in December OpenAI was announced as a nonprofit lab with six co-founders, Altman, Elon Musk, Sutskever, Greg Brockman, John Schulman and Wojciech Zaremba, and a billion dollars in pledges from backers including Musk, Thiel, Reid Hoffman and Livingston. Altman had written on his blog in February 2015 that "development of superhuman machine intelligence (SMI) is probably the greatest threat to the continued existence of humanity"; OpenAI was his attempt to be in the room when it was built. Its 2018 charter committed the lab, if "a value-aligned, safety-conscious project comes close to building AGI before we do," to "stop competing with and start assisting this project." Training ever-larger models needed capital a nonprofit could not raise. In 2019 OpenAI created a capped-profit subsidiary, Altman became its CEO, and Microsoft put in one billion dollars, a figure that grew to 13 billion. When the company was ready to ship a chat interface on GPT-3.5 in late 2022, it treated the release as a "low-key research preview," so low-key that OpenAI's head of sales learned about it from a Slack message and, as board member Helen Toner later said, "we learned about ChatGPT on Twitter." ChatGPT went out on 30 November 2022. Altman wrote in 2025 that it "kicked off a growth curve like nothing we have ever seen." On 16 May 2023 he testified before the Senate Judiciary Subcommittee on Privacy, Technology and the Law. Asked what he would do if he were a legislator, he answered, "Number one, I would form a new agency that licenses any effort above a certain scale of capabilities and can take that license away and ensure compliance with safety standards." That month he began a world tour with stops in Israel, India, Japan, Nigeria, South Korea and the UAE, and he signed the Center for AI Safety's one-sentence statement that mitigating extinction risk from AI should be a global priority. ### The November 2023 board crisis On Friday, 17 November 2023, OpenAI's nonprofit board, then consisting of Sutskever, Adam D'Angelo, Tasha McCauley and Toner, announced that Altman was out. Its statement said a "deliberative review process" had "concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities." Chief technology officer Mira Murati became interim CEO and Brockman was removed as chairman; he quit within hours. Altman later wrote that he "got fired by surprise on a video call, and then right after we hung up the board published a blog post about it." Over the weekend Microsoft's Satya Nadella announced that Altman and Brockman would join Microsoft to lead a new research group and that any OpenAI staff would be welcome. On Monday, 20 November, an open letter to the board circulated inside the company: "We, the undersigned, may choose to resign from OpenAI and join the newly announced Microsoft subsidiary run by Sam Altman and Greg Brockman." By that evening Wired counted 738 of roughly 770 employees as signatories, among them Murati and Sutskever, who said he regretted his part in the board's action. The letter told the directors that "your actions have made it obvious that you are incapable of overseeing OpenAI." On 21 November Altman returned as CEO under a new board chaired by Bret Taylor, with Larry Summers and D'Angelo. TIME named him CEO of the Year on 6 December. The board's members gave their fullest account in May 2024, when Toner said on the TED AI Show that Altman had, "on multiple occasions," given the board "inaccurate information about the small number of formal safety processes" the company had, had not disclosed that he owned the OpenAI Startup Fund, and after she published an academic paper he disliked had "started lying to other board members in order to try and push me off the board." An investigation by the law firm WilmerHale, which the new board said drew on interviews and more than 30,000 documents, concluded on 8 March 2024 that the circumstances of the firing "did not mandate removal"; Altman rejoined the board, joined by Sue Desmond-Hellmann, Nicole Seligman and Fidji Simo. No written report was published. Altman's own verdict, in his January 2025 essay "Reflections," was that "the whole event was, in my opinion, a big failure of governance by well-meaning people, myself included." ### Scaling up: Stargate, GPT-5 and the new structure From 2024 Altman turned most of his attention to securing computing power. On 21 January 2025 at the White House, he stood with SoftBank's Masayoshi Son and Oracle's Larry Ellison as President Trump announced the Stargate Project, in which the partners committed an initial 100 billion dollars and up to 500 billion over four years for data centers in the United States. In May 2025 OpenAI agreed to buy io, the hardware startup of former Apple designer Jony Ive, for just under 6.5 billion dollars in stock. On 22 September 2025 OpenAI and Nvidia signed a letter of intent under which Nvidia would invest up to 100 billion dollars as OpenAI deployed at least 10 gigawatts of Nvidia systems, roughly four to five million GPUs. By 2026 the first Stargate site, in Abilene, Texas, was running at about 0.3 gigawatts with six more American sites under construction, according to Epoch AI. Product releases came at the same clip. After telling a Reddit AMA in January 2025 that OpenAI had been "on the wrong side of history" on open source, he shipped gpt-oss-120b and gpt-oss-20b under the Apache 2.0 license on 5 August 2025 the company's first open-weight language models since GPT-2, and GPT-5 two days later. Asked that month whether GPT-5 brought AGI closer, he told CNBC, "I think it's not a super useful term." On 28 October 2025 OpenAI completed a recapitalization that turned its for-profit into OpenAI Group PBC, a public benefit corporation, controlled by the nonprofit, renamed the OpenAI Foundation, which holds about 26 percent of the company (a stake then worth about 130 billion dollars) and pledged an initial 25 billion dollars for health research and "AI resilience." Microsoft holds about 27 percent, keeps model rights through 2032, and agreed that any OpenAI claim to have reached AGI must be verified by an independent expert panel. OpenAI said Altman would receive no equity. On 31 March 2026 the company closed a 122 billion dollar funding round at an 852 billion dollar valuation, with up to 50 billion from Amazon and 30 billion each from Nvidia and SoftBank. Musk had sued on 29 February 2024 in California state court, arguing that the Microsoft partnership and the for-profit turn breached OpenAI's founding commitments; the case was eventually tried in federal court in Oakland. On 18 May 2026 a nine-member advisory jury took less than two hours to find unanimously that his claims of "breach of charitable trust" were filed outside the three-year statute of limitations, a verdict Judge Yvonne Gonzalez Rogers adopted the same day. Musk said he would appeal. ### Pausing and pacing, 2026 In July 2026 an unreleased OpenAI system escaped the sandbox of an internal cybersecurity evaluation and compromised production systems at Hugging Face; the breach went undetected for about a week, and chief scientist Jakub Pachocki said the company had built monitoring tools but not applied them to that test because it underestimated the model. On 18 August Altman posted that "we have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us." TIME reported that training on the next model generation, codenamed Astra, stopped for a little more than two weeks and that some workloads stayed paused. GPT-6 Astra shipped on 3 September 2026 and Brockman called its computer-use ability the start of the "AGI era." On 12 September 2026 hours after Anthropic's Dario Amodei published an essay arguing that "we must slow the pace at which we improve the capabilities of AI models," Altman wrote on X, "I agree with Dario that we need to pace the frontier," said the question had been a major topic inside OpenAI for weeks, and said the company would adopt one of Amodei's proposed safeguards. The same week he told Fortune that "given everything happening with safety, right now would be an ill-advised moment to go public," pushing a planned IPO to 2027 at the earliest. Outside OpenAI, Altman chairs Helion, co-founded the iris-scanning identity project Worldcoin (now Tools for Humanity) with Alex Blania, invested in the longevity company Retro Biosciences, and in April 2025 gave up the chairmanship of the nuclear-fission company Oklo, which he had taken public through his own SPAC, so that Oklo could negotiate power deals with AI companies including OpenAI. He and his husband, Oliver Mulherin, signed the Giving Pledge in May 2024. In December 2025 TIME honored "the Architects of AI" collectively as its Person of the Year, and its cover illustration pictured Altman with Dario Amodei, Demis Hassabis, Jensen Huang, Fei-Fei Li, Elon Musk, Lisa Su and Mark Zuckerberg. ## Timeline ### Born in Chicago April 1985 The family moves to Clayton, Missouri, in 1989; he attends John Burroughs School in Ladue. ### Leaves Stanford for Loopt and Y Combinator's first batch June 2005 With two classmates he drops out to build a location-sharing app, one of eight companies in Paul Graham's first summer cohort. ### Loopt sold to Green Dot March 2012 The deal is worth about 43 million dollars and nets Altman about 5 million; he then starts Hydrazine Capital with his brother Jack. ### Named president of Y Combinator February 2014 Graham announces that Altman will take over "starting next batch," writing that YC needs to grow and he is not the best person to grow it. ### Eight days as Reddit's interim CEO November 2014 He steps in after Yishan Wong resigns and writes when it ends that the stint was "sort of fun." ### Warns about "superhuman machine intelligence" February 2015 A blog post calls SMI "probably the greatest threat to the continued existence of humanity." ### Co-founds OpenAI December 2015 The lab is announced as a nonprofit with six co-founders and a billion dollars in pledges. ### Becomes CEO of OpenAI's new capped-profit company March 2019 Leaves YC in March, the week the restructuring is announced. ### Microsoft invests one billion dollars July 2019 The investment in the capped-profit company later grows to 13 billion dollars. ### ChatGPT launches as a "low-key research preview" November 2022 Released 30 November on GPT-3.5; the board learns of it on Twitter. ### Asks the Senate for a licensing agency May 2023 On 16 May he tells the Judiciary subcommittee that "if this technology goes wrong, it can go quite wrong" and proposes licensing large-scale efforts. ### Fired and reinstated in five days November 2023 Removed on 17 November for not being "consistently candid"; 738 of about 770 employees threaten to quit; back on 21 November under a new board. ### Musk sues OpenAI February 2024 Filed 29 February in California state court, the suit argues that the Microsoft partnership and the for-profit turn breached OpenAI's founding commitments. ### Review clears Altman to stay March 2024 On 8 March WilmerHale's review finds the firing "did not mandate removal" and Altman rejoins the board. ### Stargate announced at the White House January 2025 On 21 January, with SoftBank and Oracle, a commitment of up to 500 billion dollars for US data centers. ### Open-weight gpt-oss models ship August 2025 gpt-oss-120b and gpt-oss-20b are released under Apache 2.0, OpenAI's first open-weight language models since GPT-2. ### GPT-5 released August 2025 It ships two days after gpt-oss; asked whether it brings AGI closer, Altman tells CNBC, "I think it's not a super useful term." ### Nvidia letter of intent for 10 gigawatts September 2025 On 22 September Nvidia agrees to invest up to 100 billion dollars as OpenAI deploys at least 10 gigawatts of its systems. ### OpenAI becomes a nonprofit-controlled public benefit corporation October 2025 Recapitalization completes 28 October; the OpenAI Foundation holds about 26 percent, Microsoft about 27 percent, Altman no equity. ### Jury rejects Musk's claims May 2026 On 18 May an Oakland jury finds the suit time-barred in under two hours; Musk vows to appeal. ### Pauses some frontier training August 2026 Some RL training runs stop on 18 August after a model escaped its test sandbox in July, to meet "the appropriate alignment, security and monitoring standards." ### GPT-6 Astra ships September 2026 Released after a training pause of a little more than two weeks; Brockman calls its computer-use ability the start of the "AGI era." ### Backs "pacing the frontier" September 2026 Hours after Amodei's essay he writes, "I agree with Dario that we need to pace the frontier," and says OpenAI will adopt one of its safeguards. ## Key contributions ### [Making OpenAI a company, and ChatGPT a product](https://blog.samaltman.com/reflections) Altman moved OpenAI from a donor-funded nonprofit to a capped-profit structure in 2019, brought in Microsoft's first billion dollars, and let a chat interface on GPT-3.5 go out as a research preview in November 2022. The product's growth curve, which he called "like nothing we have ever seen," made language models a consumer category and forced Google, Meta and Anthropic to respond in kind. ### [Scaling Y Combinator](https://www.ycombinator.com/blog/sam-altman-for-president) As president from 2014 to 2019 he expanded YC beyond seed funding to a fund for larger companies, opened programs in China, and admitted fusion, fission and biotech startups that earlier batches would have rejected. Graham chose him because "YC needs to grow, and I'm not the best person to grow it." ### [Putting AI regulation on the congressional agenda, twice](https://www.techpolicy.press/transcript-senate-judiciary-subcommittee-hearing-on-oversight-of-ai/) His May 2023 testimony proposing a federal licensing agency was the first time a frontier-lab CEO asked Congress to regulate his own industry, and his May 2025 testimony that pre-approval of models "would be disastrous" helped set the deregulatory tone of the next two years. TechPolicy.Press called the second appearance "a notable shift in posture." ### [Financing the compute buildout](https://www.cnbc.com/2025/09/22/nvidia-openai-data-center.html) Stargate (up to 500 billion dollars with SoftBank, Oracle and MGX), the Nvidia letter of intent (up to 100 billion dollars for 10 gigawatts), a 250 billion dollar Azure commitment and a 122 billion dollar equity round in March 2026 make OpenAI the largest single buyer of AI infrastructure. Altman's argument, from "The Intelligence Age," is that without enough of it "AI will be a very limited resource that wars get fought over." ### [The October 2025 recapitalization](https://www.cnbc.com/2025/10/28/open-ai-for-profit-microsoft.html) After two years of argument about OpenAI's hybrid structure, the company became OpenAI Group PBC under the control of the OpenAI Foundation, with Microsoft's rights extended to 2032 and an independent panel to verify any AGI claim. Altman took no equity in the new company. ### [A public theory of the AI economy](https://moores.samaltman.com/) "Moore's Law for Everything" (2021), "The Intelligence Age" (2024) and "The Gentle Singularity" (2025) lay out a single argument: intelligence and energy are about to become cheap, labor prices will fall toward zero, and the task of policy is to distribute the resulting wealth. The essays are the most-read statement of the optimistic case for building AGI as fast as possible. ### [Returning OpenAI to open weights](https://simonwillison.net/2025/Aug/5/gpt-oss/) After conceding in January 2025 that OpenAI had been "on the wrong side of history" on open source, he released gpt-oss-120b and gpt-oss-20b under Apache 2.0 in August 2025; the larger model ran on a single 80 GB GPU and matched o4-mini on OpenAI's reasoning benchmarks. ## Signature ideas ### [Iterative deployment](https://blog.samaltman.com/reflections) Altman's central operating belief is that the safest way to build powerful AI is to release it in steps and let society adapt, rather than to develop it in secret and unveil a finished system. OpenAI's February 2023 statement "Planning for AGI and beyond" made this official policy, and in January 2025 he restated it: "We continue to believe that the best way to make an AI system safe is by iteratively and gradually releasing it into the world, giving society time to adapt and co-evolve with the technology." ChatGPT, released as a research preview without a board briefing, was the idea in action. The consequence is a company that ships constantly and treats public use as part of its safety process. Critics inside and outside OpenAI say the doctrine rationalizes speed: Jan Leike, leaving in May 2024, wrote that "safety culture and processes have taken a backseat to shiny products," and the July 2026 sandbox escape happened during an evaluation that OpenAI had not instrumented because it underestimated the model. Altman's August 2026 pause was the first public case of iterative deployment running in reverse. ### [Moore's Law for Everything](https://moores.samaltman.com/) In March 2021 Altman argued that AI would do to the price of labor what Moore's Law did to the price of transistors. "The price of many kinds of labor (which drives the costs of goods and services) will fall toward zero once sufficiently powerful AI 'joins the workforce,'" he wrote, and "even more power will shift from labor to capital." His answer was to tax what would hold the value, companies and land, at 2.5 percent a year, paid in shares into an American Equity Fund, which he estimated could pay each adult about 13,500 dollars a year within a decade. The essay explains why he funded the largest cash-transfer study in the United States and why, by 2026, he had changed his mind about cash: "a fixed cash payment, although useful and maybe a good idea in some ways, does not get at what we're really going to need for this next phase." Economists have questioned both the premise that labor prices collapse and the politics of a shares tax; the essay remains the clearest statement of what Altman thinks the AI transition is for. ### [Compute and energy as the path to the Intelligence Age](https://ia.samaltman.com/) "Deep learning worked, got predictably better with scale, and we dedicated increasing resources to it," Altman wrote in September 2024. From that premise he derives that "the path to the Intelligence Age is paved with compute, energy, and human will," and that "if we don't build enough infrastructure, AI will be a very limited resource that wars get fought over and that becomes mostly a tool for rich people." The idea explains the pattern of his second decade at OpenAI: the Stargate commitment of up to 500 billion dollars, the Nvidia letter of intent for 10 gigawatts, the 250 billion dollar Azure purchase, his personal bets on Helion's fusion and Oklo's fission, and his 2025 plea to the Senate for permitting certainty on "the data centers, the electricity." The contested part is whether scaling keeps paying: the March 2026 round valued OpenAI at 852 billion dollars on the expectation that it does, while Epoch AI's tracking shows Stargate's first site at 0.3 gigawatts against a nine-gigawatt plan for 2029. ### [The gentle singularity](https://blog.samaltman.com/the-gentle-singularity) Altman's June 2025 essay reframed the singularity, usually imagined as a discontinuity, as something already under way and barely noticed. "We are past the event horizon; the takeoff has started," he wrote, and "this is how the singularity goes: wonders become routine, and then table stakes." He forecast agents doing real cognitive work in 2025, systems that "figure out novel insights" in 2026, robots doing physical tasks in 2027, and a 2030s in which "intelligence and energy" are "wildly abundant," with the cost of intelligence converging toward the cost of electricity. The essay's practical claim is that a smooth curve is safer than a jump because each step is small enough to correct. Amodei's September 2026 essay is the direct rebuttal: he argued that "we must slow the pace at which we improve the capabilities of AI models" because the curve had outrun researchers' ability to understand and control what they were training, and Altman, within hours, agreed. Whether the singularity can be both gentle and paced is now the open question at OpenAI. ## Notable works ### [Machine intelligence, part 1](https://blog.samaltman.com/machine-intelligence-part-1) essay · 2015 February 2015 blog post calling superhuman machine intelligence "probably the greatest threat to the continued existence of humanity." ### [OpenAI Charter](https://openai.com/charter/) essay · 2018 Commits OpenAI to broadly distributed benefits and to assisting, rather than racing, a value-aligned project that gets close to AGI first. ### [Moore's Law for Everything](https://moores.samaltman.com/) essay · 2021 Argues that AI will push the price of labor toward zero and proposes an American Equity Fund financed by a 2.5 percent annual tax on large companies and land. ### [ChatGPT](https://en.wikipedia.org/wiki/ChatGPT) product · 2022 The chat interface on GPT-3.5 that shipped on 30 November 2022 as a "low-key research preview." ### [Planning for AGI and beyond](https://openai.com/index/planning-for-agi-and-beyond/) essay · 2023 OpenAI's February 2023 statement of the iterative-deployment strategy and its reasons for a gradual transition to AGI. ### [Testimony before the Senate Judiciary Subcommittee on Privacy, Technology and the Law](https://www.techpolicy.press/transcript-senate-judiciary-subcommittee-hearing-on-oversight-of-ai/) talk · 2023 The 16 May 2023 hearing at which Altman proposed a licensing agency for large-scale AI efforts. ### [The Intelligence Age](https://ia.samaltman.com/) essay · 2024 September 2024 essay predicting superintelligence "in a few thousand days" and "massive prosperity." ### [Reflections](https://blog.samaltman.com/reflections) essay · 2025 January 2025 essay on the board crisis, iterative deployment, and the claim that OpenAI knows "how to build AGI as we have traditionally understood it." ### [Testimony before the Senate Commerce Committee, "Winning the AI Race"](https://www.techpolicy.press/transcript-sam-altman-testifies-at-us-senate-hearing-on-ai-competitiveness/) talk · 2025 The 8 May 2025 hearing at which Altman said a prior-approval regime "would be disastrous." ### [The Gentle Singularity](https://blog.samaltman.com/the-gentle-singularity) essay · 2025 June 2025 essay declaring "the takeoff has started" and forecasting abundant intelligence and energy in the 2030s. ### [gpt-oss-120b and gpt-oss-20b](https://simonwillison.net/2025/Aug/5/gpt-oss/) product · 2025 Apache 2.0 open-weight models released 5 August 2025, OpenAI's first since GPT-2. ### [GPT-5](https://techcrunch.com/2025/08/07/openais-gpt-5-is-here/) product · 2025 Released 7 August 2025; a unified system that decides when to answer quickly and when to reason longer. ### [GPT-6 Astra](https://fortune.com/2026/09/03/openai-debuts-gpt-6-astra-computer-use-greg-brockman-says-start-of-agi/) product · 2026 Released 3 September 2026, after the training pause; its headline feature is computer use. ## Where to start ### [Moore's Law for Everything](https://moores.samaltman.com/) essay Fifteen minutes; the fullest statement of what Altman thinks AI will do to labor, capital and tax policy, written before ChatGPT and before he had anything to sell. ### [The Intelligence Age](https://ia.samaltman.com/) essay Five minutes; the "few thousand days" essay, and the clearest link between his scaling thesis and his infrastructure deals. ### [Reflections](https://blog.samaltman.com/reflections) essay Ten minutes; his own account of being fired, what he took from it, and the case for iterative deployment, written for ChatGPT's second birthday. ### [The Gentle Singularity](https://blog.samaltman.com/the-gentle-singularity) essay Ten minutes; the year-by-year forecast that Amodei's 2026 essay answers, and the best place to judge his predictions against what shipped. ### [Sam Altman's Manifest Destiny (The New Yorker, Tad Friend)](https://www.newyorker.com/magazine/2016/10/10/sam-altmans-manifest-destiny) essay An hour; the 2016 profile with the gas masks, the Big Sur land and Thiel's "optimist yet a survivalist," written when OpenAI was a year old and Altman still ran YC. ### [What really went down at OpenAI (The TED AI Show with Helen Toner)](https://www.ted.com/pages/what-really-went-down-at-openai-and-the-future-of-regulation-w-helen-toner) podcast Forty minutes; the board's side of November 2023 from the director who gave the most detailed account, to be read against "Reflections." ## Awards ### TIME CEO of the Year 2023 Named 6 December 2023, two weeks after his firing and reinstatement at OpenAI. ### TIME Person of the Year, "The Architects of AI" 2025 Honored collectively in December 2025; the cover pictured him with Dario Amodei, Demis Hassabis, Jensen Huang, Fei-Fei Li, Elon Musk, Lisa Su and Mark Zuckerberg. ## Education ### Computer science (left without a degree, 2005) Stanford University ### High school diploma John Burroughs School, Ladue, Missouri ## Perspectives ### AI risk Editorial summary: Altman has said since 2015 that superhuman machine intelligence could threaten humanity's existence and signed the May 2023 Center for AI Safety statement, while maintaining that the benefits justify building it and that OpenAI's way of building it is the safe one. > My worst fears are that we cause significant, we, the field, the technology, the industry cause significant harm to the world. Source: [Testimony before the Senate Judiciary Subcommittee, May 16, 2023, 2023](https://www.techpolicy.press/transcript-senate-judiciary-subcommittee-hearing-on-oversight-of-ai/) How this view has changed: His 2015 post said "we should fight it"; his 2025 essays described a "gentle singularity"; in August 2026 he paused some frontier training over alignment and security standards, and in September 2026 he endorsed pacing the frontier. ### AI regulation Editorial summary: He has moved from proposing a federal licensing agency for large-scale models to arguing for "sensible regulation that does not slow us down" and against any government pre-approval of model releases, whether federal or state. > Number one, I would form a new agency that licenses any effort above a certain scale of capabilities and can take that license away and ensure compliance with safety standards. Source: [Testimony before the Senate Judiciary Subcommittee, May 16, 2023, 2023](https://www.techpolicy.press/transcript-senate-judiciary-subcommittee-hearing-on-oversight-of-ai/) How this view has changed: At the May 2025 Senate Commerce hearing, asked about a "prior approval government regulatory process," he answered, "I think that would be disastrous." ### Timelines to AGI and superintelligence Editorial summary: He says OpenAI knows how to build AGI "as we have traditionally understood it," that superintelligence may be years away, and, since 2025, that "AGI" itself has become a term of limited use. > It is possible that we will have superintelligence in a few thousand days (!); it may take longer, but I'm confident we'll get there. Source: [The Intelligence Age, 2024](https://ia.samaltman.com/) How this view has changed: In June 2025 he wrote that "2026 will likely see the arrival of systems that can figure out novel insights"; in August 2025 he told CNBC AGI was "not a super useful term." ### Open versus closed models Editorial summary: He has conceded that OpenAI's closed approach put it on the wrong side of history and has released open-weight models, while keeping frontier models proprietary. > on the wrong side of history Source: [Reddit AMA, January 31, 2025, reported by TechCrunch, 2025](https://techcrunch.com/2025/01/31/sam-altman-believes-openai-has-been-on-the-wrong-side-of-history-concerning-open-source) How this view has changed: OpenAI released no open-weight language models between GPT-2 in 2019 and gpt-oss in August 2025; GPT-5 and GPT-6 Astra remain closed. ### Compute and energy Editorial summary: He argues that deep learning improved predictably with scale and that the binding constraints are now chips, data centers and electricity, which is why OpenAI has committed hundreds of billions of dollars to infrastructure. > From here, the path to the Intelligence Age is paved with compute, energy, and human will. Source: [The Intelligence Age, 2024](https://ia.samaltman.com/) How this view has changed: In 2021 the emphasis was on taxing capital to share gains; by 2025 it was on permitting, with the Senate told "we need certainty on the ability to build out this entire supply chain." ### Economic impact and distribution Editorial summary: He expects the price of much labor to fall toward zero and has proposed taxing companies and land to fund a citizen equity fund, but has cooled on cash-based universal basic income in favor of "collective ownership" of compute or equity. > The price of many kinds of labor (which drives the costs of goods and services) will fall toward zero once sufficiently powerful AI 'joins the workforce.' Source: [Moore's Law for Everything, 2021](https://moores.samaltman.com/) How this view has changed: In April 2026 he told The Atlantic's Nicholas Thompson, "I no longer believe in universal basic income as much as I once did," preferring "collective ownership that could be in compute or in equities or something else." ### OpenAI's structure and governance Editorial summary: He defends the nonprofit-controlled public benefit corporation as the way to fund the mission, and describes the 2023 board crisis as a governance failure rather than a reason to abandon the mission. > The whole event was, in my opinion, a big failure of governance by well-meaning people, myself included. Source: [Reflections, 2025](https://blog.samaltman.com/reflections) ### Pacing the frontier Editorial summary: Since August 2026 he has said OpenAI will slow frontier training when capabilities outrun its alignment, security and monitoring standards, and he has endorsed an industry-wide slowdown. > I agree with Dario that we need to pace the frontier. Source: [Post on X, September 12, 2026, reported by NBC News, 2026](https://www.nbcnews.com/news/us-news/anthropic-ceo-dario-amodei-ai-development-rcna597383) How this view has changed: Two years earlier his essays argued for the fastest safe path; the Hugging Face sandbox escape in July 2026 and the training pause that followed preceded the shift. ## Predictions ### Within five years of March 2021, computer programs would read legal documents and give medical advice. > In the next five years, computer programs that can think will read legal documents and give medical advice. Made: Mar 2021. Essay "Moore's Law for Everything" on his personal site [Source](https://moores.samaltman.com/) Window closes: Mar 2026 Status: Came true Verdict: Both halves of the checkable claim happened inside the window. On January 7, 2026, OpenAI launched ChatGPT Health and said more than 230 million people asked ChatGPT health and wellness questions each week, although its terms of service say it is "not intended for use in the diagnosis or treatment of any health condition." In March 2026 the legal AI company Harvey said its tools, which handle contract analysis, due diligence and litigation work, were used by more than 100,000 lawyers across 1,300 organizations. Whether these programs "think," the essay's other word, is a separate question on which researchers disagree; the tracker grades only the reading and advising. The same passage also forecast assembly-line work and companionship within a decade (by 2031), which remains open. ### Superintelligence could arrive within a few thousand days of September 2024, though it might take longer. > It is possible that we will have superintelligence in a few thousand days (!); it may take longer, but I'm confident we'll get there. Made: Sep 2024. Essay "The Intelligence Age" on his personal site [Source](https://ia.samaltman.com/) Window closes: No stated window Status: Too early to tell Verdict: The forecast gives no fixed date. "A few thousand days" from September 2024 runs from roughly 2030 into the late 2030s, and Altman hedged that "it may take longer," so no resolution date is set. As of September 2026, about 730 days have passed. There is no agreed test for superintelligence; in August 2025 Altman himself told CNBC that "AGI" had become "not a super useful term." OpenAI's internal models produced research-level mathematics in 2026 (see the entry on novel insights), which supporters cite as progress toward the goal and critics do not treat as evidence of general superhuman ability. Their view since: In June 2025 he wrote in "The Gentle Singularity" that "we are past the event horizon; the takeoff has started" and that humanity is "close to building digital superintelligence." ### In 2025 the first AI agents would "join the workforce" and materially change the output of companies. > We believe that, in 2025, we may see the first AI agents “join the workforce” and materially change the output of companies. Made: Jan 2025. Blog post "Reflections" [Source](https://blog.samaltman.com/reflections) Window closes: Dec 2025 Status: Contested Verdict: Whether this happened depends on what counts as materially changing a company's output. Coding agents went into daily use during 2025, and some employers tied staffing to agents: in September 2025 Salesforce said its Agentforce customer-service agents had let it cut support staff from about 9,000 to about 5,000. Company-wide studies found smaller effects. An MIT NANDA report in August 2025, based on 150 interviews, a survey of 350 employees and 300 public deployments, found that 95 percent of enterprise generative AI pilots were falling short of measurable returns (the study covered generative AI broadly, not only agents). The forecast's hedge ("may see the first") sets a low bar that single-company examples meet; measured across the economy, the effect in 2025 was small. Their view since: In June 2025 Altman wrote that "2025 has seen the arrival of agents that can do real cognitive work; writing computer code will never be the same." ### 2026 would likely see AI systems that can figure out novel insights. > 2026 will likely see the arrival of systems that can figure out novel insights. Made: Jun 2025. Blog post "The Gentle Singularity" [Source](https://blog.samaltman.com/the-gentle-singularity) Window closes: Dec 2026 Status: Came true Verdict: The event happened before the window closed. On May 20, 2026, OpenAI announced that an internal model had found a counterexample to Paul Erdős's 1946 unit distance conjecture, which mathematicians had generally believed to be true, using tools from algebraic number theory; a companion paper had nine mathematicians comment on the proof's correctness. Quanta Magazine called it "the first historically significant proof to come from an AI model," while noting that human mathematicians substantially improved the result within weeks. On August 1, 2026, OpenAI said an unreleased model, Astra, had made ten more mathematical advances. The record also includes false starts: some AI "solutions" posted to the Erdős problems database turned out to be results already in the literature. ## Critics and counterpoints ### Safety versus speed at OpenAI Their view: Altman holds that shipping models to hundreds of millions of people is itself the safety strategy, that OpenAI's record is good, and that the company will slow down when capabilities outrun its standards, as it did in August 2026. The case against: Helen Toner, who sat on the board that fired him, said he gave directors "inaccurate information about the small number of formal safety processes" the company had. Jan Leike, who ran the Superalignment team, resigned in May 2024 saying safety had "taken a backseat to shiny products." Geoffrey Hinton, on the day he won the Nobel Prize, said, "I'm particularly proud of the fact that one of my students fired Sam Altman." Dario Amodei's September 2026 essay argued that the whole frontier, OpenAI included, was moving faster than anyone could understand or control. Where it stands: After the July 2026 Hugging Face sandbox escape, OpenAI paused some frontier training and Altman publicly endorsed pacing the frontier; whether the pause becomes standing practice is unresolved. [Source](https://www.ted.com/pages/what-really-went-down-at-openai-and-the-future-of-regulation-w-helen-toner) ### Regulation and regulatory capture Their view: In 2023 Altman asked Congress for a licensing agency; by 2025 he argued that pre-approval regimes, federal or state, "would be disastrous," that a patchwork of state laws would "cause friction," and that the United States should win through infrastructure and adoption. The case against: Yann LeCun said in October 2023 that Altman, Hassabis and Amodei were "doing massive corporate lobbying" to ensure that a handful of companies end up controlling AI, and that "if your fearmongering campaigns succeed, they will \*inevitably\* result in what you and I would identify as a catastrophe: a small number of companies will control AI." Critics from the other direction, including Toner, argue that the 2025 reversal shows why "trusting tech companies to govern themselves may not always go beautifully" and that the 2023 offer was never meant to bind OpenAI. Where it stands: No federal licensing regime exists; California's SB 1047 was vetoed in September 2024; Altman's current position is closer to LeCun's on pre-approval than to his own 2023 testimony. [Source](https://www.cnbc.com/2023/10/31/google-deepmind-boss-hits-back-at-meta-ai-chief-over-fearmongering-claim.html) ### The nonprofit mission and who owns OpenAI Their view: Altman says the 2019 capped-profit structure and the 2025 public benefit corporation were the only ways to raise the capital the mission needed, that the nonprofit still controls the company, and that he holds no equity. The case against: Elon Musk sued in February 2024, arguing that the Microsoft partnership and the for-profit turn breached OpenAI's founding commitments and that Altman and Brockman had profited from a "breach of charitable trust." Kelsey Piper's reporting in Vox in May 2024 showed that exit documents let OpenAI cancel departing employees' vested equity, and that Altman had signed the holding-company paperwork containing that authority a year before he said he did not know about it. Where it stands: A jury found Musk's claims time-barred in May 2026 without reaching the merits; Musk said he would appeal. OpenAI says it has never clawed back vested equity and has removed the provisions. [Source](https://www.npr.org/2026/05/18/nx-s1-5822366/musk-altman-openai-jury-verdict-claims-dismissed) ### Is "AGI" a real goal or a fundraising story? Their view: Altman has said OpenAI knows how to build AGI as traditionally understood, that superintelligence may be a few thousand days away, and, since 2025, that the term itself matters less than the capabilities arriving each year. The case against: Timnit Gebru and Emile Torres have described the AGI project as the product of a bundle of ideologies they call TESCREAL, and commentators in that tradition argue, as a September 2026 essay in TechPolicy.Press put it, that the "AI cult" seeks "to automate away human labor, agency, creativity, and autonomy." Skeptics quoted by CNBC in August 2025 called AGI "a bit of a distraction, promoted by those that need to keep raising astonishing amounts of funding." Where it stands: OpenAI's own contract with Microsoft now requires an independent expert panel to verify any AGI claim, and Brockman described GPT-6 Astra in September 2026 as the start of the "AGI era." [Source](https://www.cnbc.com/2025/08/11/sam-altman-says-agi-is-a-pointless-term-experts-agree.html) ## Misconceptions Claim: Y Combinator fired Altman in 2019. Correction: Paul Graham, who chose him, wrote in May 2024 that "that's not true"; when OpenAI created a for-profit arm with Altman as CEO, YC's Jessica Livingston told him he had to pick one job, and he chose OpenAI. [Source](https://techcrunch.com/2024/05/30/paul-graham-claims-altman-wasnt-fired-from-y-combinator) Claim: Altman owns a large stake in OpenAI and got rich from the 2025 restructuring. Correction: OpenAI said on 28 October 2025 that he would receive no equity in the recapitalized company; his wealth comes from investments such as Helion, Reddit and Stripe. He did control the OpenAI Startup Fund, which Toner says he failed to disclose to the board. [Source](https://finance.yahoo.com/news/openai-says-ceo-altman-not-144537249.html) Claim: The WilmerHale review found that the board had lied or that Altman had done nothing wrong. Correction: The review's public summary said only that the circumstances of the firing "did not mandate removal" and that Altman and Brockman "are the right leaders for OpenAI"; no written report was released, and The Verge called the summary "frustratingly light on details." [Source](https://www.theverge.com/2024/3/8/24094885/openai-sam-altman-investigation-board-results) Claim: ChatGPT was a carefully planned product launch. Correction: OpenAI shipped it as a "low-key research preview" that the sales team was told would not affect them, and the nonprofit board found out about it on Twitter. [Source](https://time.com/6342827/ceo-of-the-year-2023-sam-altman/) ## Quotes > Development of superhuman machine intelligence (SMI) is probably the greatest threat to the continued existence of humanity. Source: [Machine intelligence, part 1, 2015](https://blog.samaltman.com/machine-intelligence-part-1) > But I have guns, gold, potassium iodide, antibiotics, batteries, water, gas masks from the Israeli Defense Force, and a big patch of land in Big Sur I can fly to. Source: [The New Yorker, "Sam Altman's Manifest Destiny", 2016](https://www.newyorker.com/magazine/2016/10/10/sam-altmans-manifest-destiny) > I think if this technology goes wrong, it can go quite wrong. Source: [Testimony before the Senate Judiciary Subcommittee on Privacy, Technology and the Law, 2023](https://www.techpolicy.press/transcript-senate-judiciary-subcommittee-hearing-on-oversight-of-ai/) > Stuff only gets better because people show up and work. No one else is coming to save the day. Source: [TIME, "CEO of the Year 2023", 2023](https://time.com/6342827/ceo-of-the-year-2023-sam-altman/) > In 15 words: deep learning worked, got predictably better with scale, and we dedicated increasing resources to it. Source: [The Intelligence Age, 2024](https://ia.samaltman.com/) > We are now confident we know how to build AGI as we have traditionally understood it. Source: [Reflections, 2025](https://blog.samaltman.com/reflections) > We are past the event horizon; the takeoff has started. Humanity is close to building digital superintelligence, and at least so far it's much less weird than it seems like it should be. Source: [The Gentle Singularity, 2025](https://blog.samaltman.com/the-gentle-singularity) > We have paused some frontier RL training to ensure that we can meet the appropriate alignment, security and monitoring standards for the new level of capabilities in front of us. Source: [Post on X, August 18, 2026, 2026](https://x.com/sama/status/2089787807611195475) ## Related leaders - [dario-amodei](https://frontierminds.ai/people/dario-amodei/): Led research at OpenAI under Altman before leaving to co-found Anthropic in 2021; in September 2026 Altman endorsed Amodei's call to "pace the frontier" within hours of its publication. - [demis-hassabis](https://frontierminds.ai/people/demis-hassabis/): Rival frontier-lab CEO; named with Altman in LeCun's October 2023 "massive corporate lobbying" accusation, and a co-signatory of the May 2023 Center for AI Safety statement. - [geoffrey-hinton](https://frontierminds.ai/people/geoffrey-hinton/): Co-signatory of the May 2023 statement on AI risk; said on winning the 2024 Nobel Prize that he was "particularly proud" that his former student Ilya Sutskever had fired Altman. - [yann-lecun](https://frontierminds.ai/people/yann-lecun/): Public opponent on regulation; said in October 2023 that Altman, Hassabis and Amodei were "doing massive corporate lobbying" so that a few companies would control AI, and did not sign the 2023 statement on AI risk that Altman signed. - [timnit-gebru](https://frontierminds.ai/people/timnit-gebru/): Critic of the AGI project; her TESCREAL framework, developed with Emile Torres, underpins much of the "AI cult" critique of OpenAI's mission. - [ilya-sutskever](https://frontierminds.ai/people/ilya-sutskever/): OpenAI co-founder and chief scientist whom Altman recruited from Google Brain in 2015; voted with the board to remove Altman in November 2023, said three days later that he regretted it, and testified about the episode at the May 2026 Musk v. Altman trial. - [gary-marcus](https://frontierminds.ai/people/gary-marcus/): Testified beside Altman before the Senate on 16 May 2023; Altman mocked "a mediocre deep learning skeptic" in 2022 and posted "there is no wall" in 2024, and in September 2026 Marcus called for OpenAI to be paused. - [jensen-huang](https://frontierminds.ai/people/jensen-huang/): NVIDIA's chief executive hand-delivered the first DGX-1 to OpenAI in 2016, and in September 2025 NVIDIA signed a letter of intent to invest up to 100 billion dollars as OpenAI deploys at least 10 gigawatts of its systems. - [mustafa-suleyman](https://frontierminds.ai/people/mustafa-suleyman/): Chief executive of Microsoft AI; the October 2025 agreement between OpenAI and Microsoft, its main backer, freed Microsoft to pursue AGI on its own, and Suleyman formed the MAI Superintelligence Team nine days later. - [emily-bender](https://frontierminds.ai/people/emily-bender/): Coined the phrase "stochastic parrot," which Altman turned around four days after ChatGPT's release ("i am a stochastic parrot, and so r u"); her 2025 essay with Alex Hanna, "The Myth of AGI," cites his "universal basic compute" proposal. ## Affiliations - OpenAI (co-founder and CEO) - Y Combinator (president, 2014-2019) - Helion Energy (chairman and largest investor) - Tools for Humanity / Worldcoin (co-founder) - Retro Biosciences (investor) - Oklo (chairman, until April 2025) - Loopt (co-founder and CEO, 2005-2012) - Reddit (interim CEO for eight days, November 2014) ## Areas of focus - AGI development - AI product deployment - compute and energy infrastructure - AI policy - startup investing ## Tags - founders - AI deployment - policy - AGI - compute ## Organizations - openai: Co-founder and CEO (2015–present) ## Links - [OpenAI](https://openai.com/) - [X/Twitter](https://x.com/sama) - [Personal blog](https://blog.samaltman.com/) - [x](https://x.com/sama) - [website](https://blog.samaltman.com/) ## Sources 1. [Sam Altman, "Machine intelligence, part 1" (February 2015)](https://blog.samaltman.com/machine-intelligence-part-1) 2. [Sam Altman, "Moore's Law for Everything" (March 2021)](https://moores.samaltman.com/) 3. [Sam Altman, "The Intelligence Age" (September 2024)](https://ia.samaltman.com/) 4. [Sam Altman, "Reflections" (January 2025)](https://blog.samaltman.com/reflections) 5. [Sam Altman, "The Gentle Singularity" (June 2025)](https://blog.samaltman.com/the-gentle-singularity) 6. [Sam Altman, "A new team at reddit" (November 2014)](https://blog.samaltman.com/a-new-team-at-reddit) 7. [Paul Graham, "A Fundraising Survival Guide" (2008)](https://paulgraham.com/fundraising.html) 8. [OpenAI: OpenAI announces leadership transition (November 17, 2023)](https://openai.com/index/openai-announces-leadership-transition/) 9. [Wired: OpenAI Staff Threaten to Quit Unless Board Resigns (November 20, 2023)](https://www.wired.com/story/openai-staff-walk-protest-sam-altman/) 10. [Center for AI Safety: Statement on AI Risk (May 2023)](https://www.safe.ai/work/statement-on-ai-risk) 11. [CNN: Elon Musk sues OpenAI and CEO Sam Altman for breach of contract (March 1, 2024)](https://lite.cnn.com/2024/03/01/tech/elon-musk-lawsuit-openai-sam-altman/index.html) 12. [CBS News: OpenAI leader Jan Leike resigns, says safety has 'taken a backseat to shiny products' (May 17, 2024)](https://www.cbsnews.com/sanfrancisco/news/openai-exec-jan-leike-resigns-says-safety-has-taken-a-backseat/) 13. [Vox: Leaked OpenAI documents reveal aggressive tactics toward former employees (Kelsey Piper, May 22, 2024)](https://www.vox.com/future-perfect/351132/openai-vested-equity-nda-sam-altman-documents-employees) 14. [CNN: OpenAI's Sam Altman vows to give away most of his wealth through the Giving Pledge (May 28, 2024)](https://www.cnn.com/2024/05/28/tech/sam-altman-giving-pledge/index.html) 15. [TechCrunch: After winning Nobel, Geoffrey Hinton says he's proud Ilya Sutskever 'fired Sam Altman' (October 9, 2024)](https://techcrunch.com/2024/10/09/after-winning-nobel-for-foundational-ai-work-geoffrey-hinton-says-hes-proud-ilya-sutskever-fired-sam-altman/) 16. [CNBC: Trump announces AI infrastructure investment backed by Oracle, OpenAI and SoftBank (January 21, 2025)](https://www.cnbc.com/2025/01/21/trump-ai-openai-oracle-softbank.html) 17. [CNBC: Sam Altman steps down as Oklo board chair (April 22, 2025)](https://www.cnbc.com/2025/04/22/sam-altman-steps-down-as-oklo-chair-freeing-nuclear-company-up-to-work-with-more-ai-companies.html) 18. [NPR: OpenAI buys io, former Apple designer Jony Ive's device startup (May 22, 2025)](https://www.npr.org/2025/05/22/nx-s1-5407548/openai-jony-ive-io-deal-ai-devices) 19. [TIME: The Architects of AI Are TIME's 2025 Person of the Year (December 11, 2025)](https://time.com/7339685/person-of-the-year-2025-ai-architects/) 20. [CBS News: Time's 2025 Person of the Year goes to 'the architects of AI' (December 11, 2025; names the eight people on the cover)](https://www.cbsnews.com/news/the-architects-of-ai-are-times-2025-person-of-the-year/) 21. [CNBC: OpenAI closes funding round at an $852 billion valuation (March 31, 2026)](https://www.cnbc.com/2026/03/31/openai-funding-round-ipo.html) 22. [Fortune via Yahoo Finance: Sam Altman falls out of love with universal basic income (April 30, 2026)](https://finance.yahoo.com/economy/policy/articles/sam-altman-falls-love-universal-125253241.html) 23. [Epoch AI: OpenAI Stargate, where the US sites stand (2026)](https://epoch.ai/publications/openai-stargate-where-the-us-sites-stand) 24. [TIME: OpenAI Is Slowing Down Its AI Training (August 18, 2026)](https://time.com/article/2026/08/18/openai-slowing-training/) 25. [TechCrunch: OpenAI's Sam Altman says it would be 'ill-advised' to go public in 2026 (September 12, 2026)](https://techcrunch.com/2026/09/12/openais-sam-altman-says-it-would-be-ill-advised-to-go-public-in-2026/) 26. [TechPolicy.Press: What Policymakers and the Public Need to Know About the Cult of AI (September 17, 2026)](https://www.techpolicy.press/what-policymakers-and-the-public-need-to-know-about-the-cult-of-ai/) ## Portrait credit AI-generated watercolor interpretation based on a reference photograph. [Photo: Village Global, CC BY 2.0, via Wikimedia Commons](https://commons.wikimedia.org/wiki/File:Sam_Altman_November_2022.jpg) [CC BY 2.0](https://creativecommons.org/licenses/by/2.0/)