OpenAI IPO Plans Face Internal Pushback as CFO Sarah Friar Flags Risks and Leadership

OpenAI’s board is now wrestling with a stark financial forecast that could upend its IPO timetable, according to the Wall Street Journal. CFO Sarah Friar warned that the company expects to burn $85 billion in 2028 even after nearly doubling sales, a loss “that would dwarf that of virtually any other public company in history” (Wall Street Journal). The warning has triggered a surge of internal dissent, with senior engineers and product leads questioning whether the firm can sustain such a cash‑drain while courting public investors.

The cost shock stems from the company’s aggressive model‑training agenda. The Journal’s deep‑dive shows OpenAI plans to spend $121 billion on computing power for AI research by 2028, far outpacing rival Anthropic’s projected outlays (Wall Street Journal). Those figures come from confidential financial decks shared with investors earlier this year and echo earlier reporting by The Information on the same data. The expense surge reflects the “arms race” to release new model versions at a faster cadence, a strategy that demands ever‑larger GPU clusters and custom silicon (Wall Street Journal).

Board members are now confronting a leadership‑structure dilemma. Friar’s memo flagged not only the fiscal risk but also a “potential misalignment” between the executive team and the research org, which she says could impair cost controls (Wall Street Journal). Sources at the company, speaking on condition of anonymity, said the CFO’s concerns have prompted a review of the current reporting lines, with some senior managers pushing for a tighter integration of finance and engineering to curb runaway spending (Republic World). The internal pushback has reportedly slowed the finalization of the IPO filing schedule, which had been slated for the fourth quarter.

Investors are watching the fallout closely. The Journal notes that the projected losses would be unprecedented for a public tech firm, raising questions about valuation metrics that have buoyed OpenAI’s $157 billion market cap in private rounds. Analysts cited by Republic World warned that any perception of uncontrolled cost growth could depress demand for shares once the company goes public. Meanwhile, OpenAI’s rivals, including Anthropic, are also grappling with rising compute bills, though at a lower scale, suggesting the sector-wide pressure may temper market enthusiasm (Wall Street Journal).

The company has not ruled out adjusting its IPO timeline. Friar’s internal briefing, as reported by the Wall Street Journal, indicates the board may delay the offering until a clearer path to profitability emerges. Executives are reportedly exploring cost‑reduction measures such as scaling back the frequency of major model releases and renegotiating cloud‑provider contracts (Republic World). As the internal debate intensifies, OpenAI’s next public move will hinge on whether it can align its ambitious research agenda with the fiscal discipline demanded by public markets.