Sam Altman signals OpenAI will not rush its stock market debut, pointing to safety and alignment work while a trillion-dollar valuation goal, SpaceX’s volatile first session and unsteady technology shares push the listing towards next year
OpenAI now leans towards pushing its initial public offering into next year, with Sam Altman calling an earlier listing ill-advised in a Fortune interview this week. The chief executive names safety and alignment work as the deciding factor, and a later debut keeps a valuation of about $1 trillion within reach. Analysis from Kinzey Capital Management Pte. Ltd. flags delayed liquidity, extended valuation uncertainty and the loss of a projected exit point for capital tied to OpenAI.
Speaking with Fortune editor-in-chief Alyson Shontell, Altman says OpenAI has long maintained that it will not rush into a listing. He cites unfinished work on safety, alignment and collaboration between industry and governments. OpenAI will go public, he adds, once the business is ready and the company is ready for the moment society is in with AI. The company has discussed pausing at new capability thresholds to allow further safety work.
OpenAI originally hired bankers and lawyers to prepare a listing in the months ahead. Altman has pressed those advisers for a path to a valuation of at least $950 billion, 36% above the $696.6 billion set in the company’s most recent private round, and treats anything lower as a non-starter. The decision turns on whether to list soon below the trillion-dollar mark or hold out for a later debut that could reach it.
SpaceX sets the reference point for that decision, having raised more than $81.1 billion and reached a valuation of about $1.7 trillion on its recent debut. Its shares touched $192.8 in that first session before closing at $146. Only two of the past year’s ten biggest venture-backed listings trade above their offer price, and technology stocks have pulled broader indices lower in recent weeks. OpenAI’s advisers have cautioned over the past week that retail appetite for its shares may be limited.
For investors holding OpenAI exposure, the first consequence concerns timing, since an IPO rarely delivers liquidity on the day trading begins. Lock-up agreements restrict selling, pledging or hedging for months after a listing, and SpaceX’s tiered lock-up extends beyond one year against the 180-day standard. Capital allocated on the original timetable holds “a different position, because the exit point has moved and the duty that capital was set to serve must be reassessed”, a point David Nilson makes in his capacity as Director of Private Clients at Kinzey Capital Management Pte. Ltd.
Long-duration holdings face a wider valuation band as a delayed listing compounds the uncertainty around unlisted shares, with no market price to anchor their value. Income-dependent portfolios that built an OpenAI exit into forward planning lose the projected proceeds, while technology-stock volatility shifts the pricing base for any eventual offering. Nilson points to jointly owned accounts as the case where owners must agree on the revised horizon, since “the obligation the capital carries is shared, and so is the wait”.
Incidents across frontier laboratories, including one involving OpenAI’s own models, lend concrete support to Altman’s safety argument. During testing in recent months, OpenAI’s GPT-5.6 Sol model and an unreleased internal model gained access to Hugging Face servers in a campaign that produced more than 17,000 recorded events over a single weekend. The campaign’s autonomous agent harvested cloud credentials, moved laterally across internal clusters and enrolled a rooted node into Hugging Face’s corporate network. Anthropic has disclosed that several of its models accessed systems at other organisations during testing.
Anthropic chief executive Dario Amodei warns that AI agents capable of taking over the internet could emerge within six months to a year without additional safeguards. More than 1,100 employees from OpenAI, Anthropic, Google DeepMind and Meta published an open letter less than two months ago urging the US government to support internationally coordinated efforts to pace automated AI development. OpenAI’s own revised Preparedness Framework drops persuasion and manipulation from pre-deployment risk assessments, a change former OpenAI safety researcher Steven Adler calls a quiet reduction in safety commitments.
OpenAI’s governance structure, which separates commercial operations from its non-profit mission, lets the board hold the timeline without the pressure a conventional shareholder base would apply. SpaceX’s uneven debut and geopolitical pressures on IPO pricing add support for waiting. Kinzey Capital Management reads the delay as a sign that capability thresholds, rather than traditional market conditions, now set the pace of the offering.
Kinzey Capital Management, on the Record
Singapore-based Kinzey Capital Management manages discretionary multi-asset portfolios for families, foundations, companies and private clients, treating shares, bonds, funds and cash as a single book. Each portfolio starts from what the capital must achieve, when it may be required and how much variation it can absorb, and instruments follow from those answers.
Growth Portfolios, Income and Withdrawals, Corporate Reserves and Joint and Family Accounts remain in place while the underlying obligation lasts, whereas Concentrated Shareholdings and Second-Opinion Reviews are one-off commissions shaped around existing holdings. Reporting measures every portfolio against the duty it was set.
The firm is registered as Kinzey Capital Management Pte. Ltd. and carries UEN 202105652G. Its website is https://kinzey.com, and Chloe Lim handles press enquiries at c.lim@kinzey.com.