According to monitoring by Dongcha Beating, The Wall Street Journal published an in-depth investigation revealing an internal deadlock at OpenAI on the eve of its IPO: CEO Sam Altman proposed that OpenAI invest $500 million in a new funding round for the fusion startup Helion, but employees rejected the deal after evaluation. This is also one of the reasons why some shareholders have begun privately discussing the possibility of replacing Altman as CEO, with board chairman Bret Taylor (former co-CEO of Salesforce) potentially taking over. Helion is Altman's most significant bet outside of OpenAI, and he is one of the company's largest investors, with a substantial portion of his personal net worth tied to it; in 2021, he made a one-time investment of $375 million, which was his largest personal investment at that time. The funding round was originally planned to be around $1 billion, with a valuation of $35 billion, while the company's valuation was only $5.4 billion during its last funding round in January 2025. Employees at OpenAI have two main concerns regarding the proposal. First, Altman is asking OpenAI to invest in a company where he holds a significant stake, which ties a considerable amount of his net worth, while OpenAI itself would not directly benefit from the business. Second, employees are skeptical about Helion's fusion technology; the company promised to generate more power than it consumes with its Polaris device by 2024 but missed that deadline and has since only vaguely stated that 'the machine has achieved several technical milestones' without providing specific data. Insiders say some employees deliberately avoid discussing the deal in Slack channels, fearing that their comments could later be summoned in legal proceedings. Although OpenAI rejected the investment, it signed a power purchase agreement that allows it to procure up to 50 gigawatts of power from Helion by 2035, equivalent to the output of 25 Hoover Dams. Helion subsequently used this contract as a selling point for the new funding round, but the financing target has been reduced to $250 million, with a valuation of $15 billion, now led by Thrive Capital (one of OpenAI's major shareholders). Altman stepped down from his board position at Helion last month. The entanglement of Altman’s interests with OpenAI also extends to space. The WSJ first disclosed that Altman and his husband hold shares in the rocket company Stoke Space through their family office, Hydrazine. Last summer, he proposed that OpenAI acquire or become a majority shareholder in Stoke to build data centers in space, directly challenging Elon Musk's SpaceX. After the WSJ began inquiring about this matter in December, related negotiations were paused; Altman publicly called the idea of space data centers 'absurd' during an event in India in February, surprising those involved in the negotiations. Insiders say Altman is still privately pushing for a rocket launch cooperation agreement this year. Internal personnel matters are also delicate. Chief Product Officer Fidji Simo was originally expected by Altman to take over most daily operations after the IPO, including representing the company at quarterly earnings meetings, but she announced a medical leave this month due to a relapse of a neuroimmune disease, designating four executives to share her responsibilities, with Altman not on the list. Taylor, however, expressed support for Altman in a statement, saying he 'sees every day why Sam is uniquely suited to lead the company into the next phase.' Behind these conflicts lies a structural root: OpenAI was spun off from a non-profit organization, and Altman currently has no direct equity in the company, with a salary of only $66,000 in 2024, while his wealth is highly tied to a portfolio of hundreds of startups accumulated during the YC era. Whenever he asks OpenAI to take action, it inevitably raises the question of 'is this to support his own holdings,' and his subordinates are now beginning to respond to this with legal risks.