OpenAI CFO Sarah Friar Privately Pushing for 2027 IPO Delay — Internal Dispute With Altman

Q: What this means for OpenAI's trajectory

Three implications. First, expect continued public messaging discipline from OpenAI about IPO timing — neither 2026 nor 2027 will be officially confirmed until the decision is fully made and the SEC filing is ready. Second, expect SoftBank and other large private backers to play a meaningful role in the timing decision, given their capital exposure and their strategic interests in OpenAI's path to liquidity. Third, expect broader AI-IPO calendar to shift — a 2027 OpenAI listing would push other AI companies (Anthropic, possibly Perplexity) to align their timing accordingly to avoid being eclipsed by OpenAI's debut. For private-market valuations, the IPO timing matters less than the structural maturation it represents. Whether OpenAI lists in early 2026 or 2027, the multi-billion-dollar capital flows in and out of frontier AI will continue at roughly the current pace. The IPO will be a major optical event but is unlikely to fundamentally shift the underlying capital dynamics of the AI industry.

By Jaspal Singh May 3, 2026 Updated: May 4, 2026

OpenAI CFO Sarah Friar has privately suggested delaying the company's much-anticipated IPO to 2027, the Wall Street Journal reported yesterday, citing people familiar with the discussions. The push for delay reflects ongoing tension between Friar's preference for a more conservative listing timeline that lets the company demonstrate sustained revenue growth, and CEO Sam Altman's reported preference for moving faster on the public-market debut. The internal disagreement is the most consequential governance dispute at OpenAI since the November 2023 board crisis.

The strategic context matters. OpenAI has been widely expected to file for IPO in late 2026 with a target listing in early-to-mid 2026, at a valuation reportedly in the $400–500 billion range. Friar — who joined OpenAI in mid-2024 from Nextdoor and Square — has reportedly been arguing that pushing the listing to 2027 would let OpenAI hit several growth and profitability milestones that would meaningfully strengthen the public-market thesis. The disagreement is fundamentally about whether OpenAI should optimize for the highest possible valuation at IPO or for a more durable post-listing trajectory.

Why Friar wants to wait

Three specific concerns appear to be driving Friar's preference. First, profitability optics: OpenAI is currently spending materially more on training compute than it's generating in revenue, and the gap is widening as the company invests in next-generation models. Going public while operating losses are still expanding creates a difficult narrative for institutional investors who increasingly demand path-to-profitability metrics from late-stage AI listings. A 2027 timeline would let OpenAI demonstrate that operating leverage is improving — or at least stabilizing.

Second, revenue mix concentration: ChatGPT consumer subscriptions and a small handful of enterprise contracts (including the Microsoft revenue-share arrangement) are still the dominant revenue lines. Friar has reportedly emphasized building broader enterprise diversification before public-market scrutiny intensifies, and that takes time. Third, regulatory uncertainty: the EU AI Act enforcement phase, evolving U.S. AI export controls, and ongoing copyright litigation all create disclosure complications that get harder to manage in public-company SEC filings. Resolution of those issues — or at least more clarity on their trajectory — would substantially de-risk the IPO.

Why Altman wants to move faster

The reported counter-argument from Altman emphasizes capital and competitive dynamics. OpenAI needs significant ongoing capital for its training-compute commitments and infrastructure ambitions, and a public listing creates the most efficient long-term capital-raise mechanism. Continuing to fund operations through massive private rounds (the $40B+ commitment from SoftBank-Stargate, follow-ons from existing backers, etc.) is workable but increases the company's dependence on a small set of large institutional partners with their own strategic interests.

The competitive dimension also matters. Anthropic is reportedly raising at a $900B valuation, and the gap between OpenAI's last private valuation ($300B) and Anthropic's prospective valuation has materially narrowed. Going public earlier, at a stronger valuation, would re-establish OpenAI's premium positioning and create a higher-water-mark anchor that competitors would have to chase. Waiting to 2027 risks letting Anthropic's valuation creep level the optical playing field.

My Take

Friar is probably correct on the merits. Public-market investors in 2026 are increasingly skeptical of AI companies that haven't shown clear path to profitability, and OpenAI's current operating-loss profile would create immediate post-IPO pressure if the listing happened before margin trajectory improves. The Stripe and Klarna IPOs of 2025 both demonstrated that even well-loved private companies can see meaningful post-IPO underperformance when the operating-margin story isn't ready.

The deeper governance question is whether OpenAI's structure — with the for-profit subsidiary controlled by a non-profit board with capped financial returns — is even compatible with traditional public-market expectations. The conversion to a more standard for-profit structure has been telegraphed but not formally completed, and IPO timing is inseparable from that structural overhaul. Pushing to 2027 effectively buys time for both the financial maturation and the corporate-structure resolution to happen in parallel, which is materially cleaner than trying to do both during an active IPO process.

For Altman specifically, the IPO timing dispute is also tactically important because a successful public listing locks in his position as the public face of OpenAI in a way that protects against any future board interventions. The November 2023 crisis demonstrated that even a CEO with strong external visibility can be removed by a non-profit board acting on its own counsel. An IPO and the resulting public shareholder accountability would effectively immunize the CEO position from that risk going forward. That's a meaningful personal and institutional consideration that probably shapes Altman's preference for sooner.

What this means for OpenAI's trajectory

Three implications. First, expect continued public messaging discipline from OpenAI about IPO timing — neither 2026 nor 2027 will be officially confirmed until the decision is fully made and the SEC filing is ready. Second, expect SoftBank and other large private backers to play a meaningful role in the timing decision, given their capital exposure and their strategic interests in OpenAI's path to liquidity. Third, expect broader AI-IPO calendar to shift — a 2027 OpenAI listing would push other AI companies (Anthropic, possibly Perplexity) to align their timing accordingly to avoid being eclipsed by OpenAI's debut.

For private-market valuations, the IPO timing matters less than the structural maturation it represents. Whether OpenAI lists in early 2026 or 2027, the multi-billion-dollar capital flows in and out of frontier AI will continue at roughly the current pace. The IPO will be a major optical event but is unlikely to fundamentally shift the underlying capital dynamics of the AI industry.

Frequently Asked Questions

When will OpenAI actually IPO?
Uncertain. Public reporting suggests CFO Sarah Friar is privately advocating for a 2027 listing, while CEO Sam Altman reportedly prefers a faster 2026 timeline. The decision is still being made internally. Expect formal disclosure within 6–9 months.

What's OpenAI's current valuation?
The most recent disclosed private valuation was approximately $300 billion. Public reporting suggests OpenAI could target a $400–500 billion valuation at IPO depending on timing and market conditions.

Why are people talking about a 2027 delay?
The Wall Street Journal reported yesterday, citing people familiar with internal discussions, that CFO Sarah Friar has been advocating for a delay. The reporting describes the disagreement as a meaningful governance dispute between Friar and CEO Sam Altman.

Does this affect the OpenAI corporate structure?
Indirectly. OpenAI's transition from a non-profit-controlled for-profit subsidiary to a more standard public-market structure is closely tied to IPO timing. A 2027 listing gives more time for the structural overhaul to complete cleanly.

The Bottom Line

OpenAI's IPO timing dispute between CFO Friar (2027) and CEO Altman (faster) is the most consequential governance discussion at the company since 2023. Friar's argument for a 2027 timeline is well-grounded in public-market realities, but Altman's institutional and tactical motivations for moving faster are also coherent. Watch for resolution within 6–9 months — and expect the eventual decision to materially shape the broader AI-IPO calendar.