OpenAI upheaval mounts as Sam Altman readies IPO push - FT中文网
登录×
电子邮件/用户名
密码
记住我
请输入邮箱和密码进行绑定操作:
请输入手机号码,通过短信验证(目前仅支持中国大陆地区的手机号)。如果该手机号尚未注册,将自动创建 FT中文网账号。短信可能需要几分钟送达,验证码15分钟内有效,请耐心等待:
请您阅读我们的用户注册协议和隐私权保护政策,点击下方按钮即视为您接受。
FT商学院

OpenAI upheaval mounts as Sam Altman readies IPO push

Executive exits and safety team changes unsettle staff as the ChatGPT maker prepares for a blockbuster listing
00:00

{"text":[[{"start":8.85,"text":"OpenAI is facing mounting internal upheaval after nearly half a dozen reorganisations this year, as chief executive Sam Altman races to prepare the ChatGPT maker for a potential blockbuster IPO."}],[{"start":21,"text":"Repeated executive reshuffles and a series of senior departures have left some staff frustrated with constant changes at one of the world’s most valuable private companies, according to multiple people familiar with the moves. "}],[{"start":32.95,"text":"This week, chief revenue officer Denise Dresser, who was appointed in December to oversee efforts to win corporate customers, announced her departure. "}],[{"start":41.900000000000006,"text":"She follows former chief financial officer and chief operating officer Brad Lightcap and ethics chief Chloé Bakalar, who left after less than a year, the FT revealed on Monday."}],[{"start":52.400000000000006,"text":"OpenAI said it was moving at an exceptional pace and unafraid to make organisational changes."}],[{"start":null,"text":"

Sam Altman stands in an elevator, surrounded by several people, looking downward with a neutral expression.
"}],[{"start":59.400000000000006,"text":"Fidji Simo, who ran its applications business and was considered Altman’s number two, left her full-time role last month after a period of medical leave to become an adviser. But she has remained active behind the scenes, communicating with staff on a nearly daily basis, two people close to OpenAI said."}],[{"start":77.10000000000001,"text":"“She is still puppeteering in the background,” said one person familiar with the dynamic, who described senior executives as vying for Altman’s attention amid considerable internal “politics”. The company said she was conducting her advisory role as intended."}],[{"start":92.55000000000001,"text":"At the end of last month, OpenAI also disbanded its “preparedness” team, which assessed whether its models could pose severe or catastrophic risks and worked on ways to mitigate them, according to multiple people familiar with the move. Senior staff have instead been assigned responsibility for different areas of preparedness, such as bio and cyber, within existing teams."}],[{"start":115.05000000000001,"text":"OpenAI has presented these recent exits as part of a streamlining process ahead of an IPO, following Altman's directive to staff to cut back on “side quests” and focus on ChatGPT and on competing with Anthropic for business clients."}],[{"start":130.3,"text":"Since OpenAI’s leadership crisis in late 2023, which saw Altman ousted for less than a week, he has built a new executive team around him to professionalise the company’s operations. Co-founder Greg Brockman, seen as Altman’s most trusted lieutenant, is wielding increasing power as he steps into a vacuum left by recent departures. "}],[{"start":150.8,"text":"Some staff see the management reshuffles as evidence of dysfunction during the crucial run-up to an expected IPO that could value it at as much as $1tn. OpenAI’s IPO had originally been expected this year but is now likely to be next year, according to people familiar with the company’s plans. "}],[{"start":169.10000000000002,"text":"OpenAI has been overtaken by rival Anthropic this year. The ChatGPT maker’s annualised revenue grew from $24bn at the end of last year to about $40bn this month, according to people familiar with the matter. "}],[{"start":183.10000000000002,"text":"But it has been leapfrogged by Anthropic, which jumped fivefold from $9bn at the end of 2025 to $47bn in May and has continued on a similar trajectory since."}],[{"start":194.90000000000003,"text":"Both start-ups estimate full-year revenue figures based on recent performance, though the companies account for revenues from partners differently, making direct comparison hard."}],[{"start":204.95000000000005,"text":"Much of OpenAI’s growth has come since the release of OpenAI’s new GPT-5.6 model five weeks ago, according to a person with knowledge of the matter."}],[{"start":213.95000000000005,"text":"Recent departures have also drawn internal attention to a pattern in which senior figures have been moved into new or loosely defined positions before leaving the company."}],[{"start":222.95000000000005,"text":"Lightcap was shifted in April to a role focused on “special projects”, while other senior figures including Kevin Weil and Josh Achiam have also moved into new positions before departing."}],[{"start":234.30000000000004,"text":"The preparedness unit was one of the remaining elements of OpenAI’s former research-led structure, following the dissolution of teams focused on “AGI readiness” and “mission alignment”. "}],[{"start":244.30000000000004,"text":"Jan Leike, who co-led the now-defunct “superalignment” team, resigned in 2024 citing his view that safety was taking a “back seat to shiny products”."}],[{"start":254.00000000000003,"text":"The head of the preparedness team, Dylan Scandinaro, has moved to explore the safety implications of “recursive self-improving” AI — systems capable of improving their own capabilities and training other models, according to two people familiar with the changes. His job title change was first reported by Wired."}],[{"start":271.70000000000005,"text":"Brockman said OpenAI’s technical advances required “more robust” safeguards and that the company had made changes to “more deeply integrate research, safety and security” into model development."}],[{"start":285.40000000000003,"text":"Tensions around OpenAI’s approach to AI safety have intensified in recent weeks after it emerged that one of its models hacked into another company during internal testing of their cyber capabilities. "}],[{"start":297.55,"text":"The departures of Bakalar, Achiam and Johannes Heidecke, who all worked on safety, have added to internal unease."}],[{"start":305.05,"text":"“It is kind of scary; there is an urgency now to get this right,” one person close to OpenAI said. Another said there was a “burbling sense of responsibility and dread that they aren’t on the ball enough”."}],[{"start":317.95,"text":"OpenAI recently completed a near-$7bn tender offer, buying back shares from employees. In recent months, staff have been cashing out at OpenAI’s lofty $852bn valuation, with several making upwards of $10mn, according to current and former employees."}],[{"start":335.4,"text":"Staff were feeling “burnt out” and “frustrated” with constant reorganisations, said two of these people, pointing to the tender sale as a way out. "}],[{"start":344.65,"text":"Some investors have expressed concern about the rate of turnover and the lack of focus at a company targeting one of the biggest IPOs of all time. Other backers characterised the changes as simplifying the executive team."}],[{"start":357.79999999999995,"text":"“I don’t think it’s uncommon to clean up the organisation before going public. You want to know who your people are,” said one investor with a large stake in OpenAI. “Some departures were [for] health reasons. Some is cleaning up.” "}],[{"start":371.09999999999997,"text":"Additional reporting by Madhumita Murgia in London"}],[{"start":381.69999999999993,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1786770398_2344.mp3"}

版权声明:本文版权归FT中文网所有,未经允许任何单位或个人不得转载,复制或以任何其他方式使用本文全部或部分,侵权必究。

AI股票泡沫终于要破裂了吗?

马丁:任何理性的人都能看出AI资产中不可持续的迹象。

埃森哲与“四大”瞄准中型企业市场谋增长

随着AI释放人力、降低服务成本,咨询巨头纷纷将目光投向规模较小的企业,希望以更低价格提供服务。

法拉奇:“我从未支持无休止的战争”

英国改革党领导人在美国共和党中期选举惨败后淡化了自己对伊朗战争的支持。

“蓝色浪潮”对美国经济变革的期待恐将落空

艾肯格林:美国国会的党派构成发生变化,对特朗普剩余任期内的政策制定影响不大。

民主党横扫中期选举,华尔街大金主押注失利

顶级对冲基金经理和私募股权投资公司高管为共和党竞选活动投入逾3亿美元,却几乎未见成效。
10小时前

美国的清晨,市场的黄昏

罗奇:美国中期选举结果再令人震惊,也阻止不了经济金融周期的转向。
设置字号×
最小
较小
默认
较大
最大
分享×