Tools chosen: kalshi_direct kalshi_related polymarket_related claude_news claude_news gdelt_news wikipedia code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a relative-timing question between two private AI labs, so the key drivers are reported IPO preparation signals (bankers, CFO hires, restructuring) for each company. The Kalshi price is the primary anchor, and recent news (OpenAI's late-2025 for-profit restructuring and reported 2026-2027 IPO groundwork vs. Anthropic's reported IPO exploration) will determine whether to deviate.

# Current state Kalshi YES (Anthropic-first) is at 84%, up from a 90-day low of 22% and near its 90-day high of 88%. Both OpenAI and Anthropic have confidentially filed S-1s with the SEC (Anthropic June 1, 2026; OpenAI ~May 22, 2026), but neither has held a public roadshow or set a public listing date. Anthropic filed slightly before OpenAI and is reported to be further along in visible IPO-readiness steps (investor meetings, credit lines, CFO-led process), which appears to be driving the price toward Anthropic. # Timeline of key events - 2025-Q4: OpenAI completes for-profit PBC restructuring ("OpenAI Group PBC"), 26% owned by nonprofit OpenAI Foundation; valuation ~$500B (confirmed, finance.yahoo.com/seekingalpha.com). - 2025-11: OpenAI CFO Sarah Friar says "IPO is not on the cards right now" (confirmed quote, finance.yahoo.com). - 2026-01/02: Anthropic closes $30B Series G at $380B valuation (reported, CryptoBriefing). - 2026-04-29: Reports (CNBC/Bloomberg) of pending $50B Anthropic round that would surpass OpenAI's valuation; board decision expected May, potential IPO "as early as October 2026" per TechCrunch (reported). - 2026-04-08: Friar tells CNBC OpenAI tested retail demand in latest funding round, "really strong demand" (confirmed quote). - 2026-05-06: Anthropic implied pre-IPO valuation hits $1.2T after 20% 7-day surge (reported, Economic Times). - 2026-05-20/22: OpenAI confidentially files draft S-1 with SEC (reported/confirmed across multiple outlets, valuation ~$852B post-funding round). - 2026-06-01: Anthropic announces it filed confidential S-1, closes $65B round at $965B valuation, surpassing OpenAI (confirmed, finance.yahoo.com, Al Jazeera). - 2026-06-02: Wedbush (Dan Ives) calls Anthropic's filing "a major step... to get ahead of OpenAI" (analyst opinion). - 2026-06-09: Altman reportedly tells staff OpenAI could go public "within the next year"; wants a September listing but no decision on timing (reported, Fortune/TS2). - 2026-06-26: CNBC/Kalshi reporting: OpenAI IPO timeline seen slipping toward 2027; Kalshi (referenced market, not this ticker) implies ~59% chance of announcement by March 2027 (reported). - 2026-07-15/17/21: Anthropic reportedly pursuing multibillion-dollar pre-IPO credit lines (SpaceX-style) and scheduling investor meetings ahead of potential IPO (reported, multiple outlets). - 2026-07-29: Friar tells employees OpenAI's July annualized revenue topped all of Q2 (confirmed CNBC). - Late July 2026 (latest data point): Both firms have confidentially filed S-1s; neither has disclosed a public timeline (confirmed, CNBC). # Event Will Anthropic confirm an IPO before OpenAI (mutually exclusive first-mover market, resolving by 2040)? # Outcomes to forecast - Yes (Anthropic IPOs first) - No (OpenAI IPOs first, or Anthropic does not IPO first) # Kalshi market anchor KXOAIANTH-40-ANTH currently trades at 84% YES (Anthropic first), +3% over 7 days and +3% over 30 days. 90-day range is wide (22%–88%), average daily volume ~1,169 contracts — indicates a large repricing has already occurred (from ~22% low up to 84-88%), likely coinciding with Anthropic's June 1 filing news and subsequent Q2 profitability/valuation reports. This is the primary consensus to beat. # Sub-question answers 1. **Current Kalshi price/90-day trend** — 84% YES, up from a 22% low over 90 days to an 88% high recently; trending up (+3%/7d, +3%/30d). [kalshi_direct] 2. **OpenAI IPO-prep steps** — Completed PBC conversion (Oct 2025, $500B valuation, Microsoft deal); confidentially filed S-1 ~May 22, 2026; CFO Friar initially resistant (Nov 2025: "not on the cards"), later softened (Apr 2026: testing retail demand); Altman reportedly wants Sept 2026 listing but no committed date; reports (June 2026) suggest slippage toward 2027. [claude_news, gdelt_news] 3. **Anthropic IPO-prep steps** — CFO Krishna Rao (ex-Airbnb IPO exec) steering process; confidentially filed S-1 June 1, 2026, ahead of OpenAI; closed $65B round at $965B valuation same period; pursuing multibillion-dollar pre-IPO credit lines and investor meetings (July 2026); targeting October 2026 per multiple reports. [claude_news, gdelt_news] 4. **Structural obstacles** — OpenAI's nonprofit Foundation retains 26% control post-restructuring but the PBC conversion (completed) removes the primary structural IPO blocker; Anthropic is a PBC with no reported blocking trust structure, though a Pentagon legal dispute is cited as a complicating (non-structural) distraction for its CFO. [wikipedia, claude_news] 5. **Valuations/cash needs** — Anthropic: $965B (June 2026, some reports to $1.2T implied); reportedly near-profitable (~$559M Q2 operating income on $10.9B revenue). OpenAI: $852B (early 2026 round); reportedly still loss-making, expected to burn >$200B before cash-flow positive — creating potential pressure for earlier listing despite caution. [claude_news] 6. **Parallel markets** — No matching Polymarket markets found (0 hits). Other Kalshi series (bank-lead-OpenAI-IPO, Ramp/Brex, Deel/Rippling) show comparable "first mover" market structures but no direct cross-check on Anthropic-vs-OpenAI odds elsewhere; one referenced (non-ticker) Kalshi data point cited by BitMEX blog claims "72% chance Anthropic lists before OpenAI" — roughly consistent with current 84% anchor's directionality, though not identical timeframe/market. [kalshi_related, claude_news] # Key facts (high-confidence, factual) 1. [wikipedia] Both firms are PBCs; OpenAI's nonprofit Foundation holds 26% of OpenAI Group PBC. 2. [finance.yahoo.com] Anthropic confidentially filed S-1 June 1, 2026, at $965B valuation. 3. [multiple, confirmed] OpenAI confidentially filed draft S-1 ~May 22, 2026, at $852B valuation. 4. [cnbc.com, late July 2026] Neither company has disclosed a public listing timeline as of the latest reporting. 5. [finance.yahoo.com] OpenAI's PBC conversion (Oct 2025) removed nonprofit-only structural block to IPO eligibility. # Cross-market signals - Kalshi related: Bank-lead-OpenAI-IPO market (Goldman Sachs favorite, 59%, declining) implies market still actively pricing OpenAI IPO mechanics, not doubting it will happen. - Polymarket: No relevant markets found. - Other cited (non-verified) Kalshi data via BitMEX blog: "72% Anthropic-first" — directionally consistent with current 84% anchor. # Analyst opinions and speculation - Wedbush (Dan Ives): Anthropic's filing is "a major step... to get ahead of OpenAI." [Fortune] - NYT/CNBC (reported, not confirmed): OpenAI may delay to 2027 amid Altman-Friar internal disagreement. - Code-execution Monte Carlo model (base case) estimated fair YES ~19-20%, far below current 84% market price — but this model relied on stale/generic timing assumptions (OpenAI 2027.5, Anthropic 2029) that pre-date the actual June 2026 filing news; treat as outdated and low-weight versus the news-driven Kalshi anchor. # Directional lean per outcome - **Yes (Anthropic first)**: Filed S-1 first (June 1 vs ~May 22, but announced/publicized as "beating" OpenAI); reportedly profitable; more concrete recent prep steps (credit lines, investor meetings) surfacing in July 2026; market has re-priced sharply upward to 84%. - **No (OpenAI first)**: OpenAI filed its confidential S-1 slightly earlier in calendar terms; Altman wants speed (Sept 2026 target); OpenAI has stronger revenue-growth headlines (July ARR surpassing Q2). But internal caution (Friar) and reported 2027 slippage weigh against. # Gaps / unknowns - No confirmed public S-1 (only confidential filings for both) — actual listing dates unknown. - Unclear which SEC filing legally counts as "first" vs. which company completes public listing first (market resolves on confirmed IPO, not filing). - Pentagon lawsuit's material impact on Anthropic's timeline is unquantified. - No Polymarket cross-check available. # Calibration anchors - Kalshi current YES price: 84% (anchor). - 90-day range 22%–88% shows large, recent repricing following June 2026 filing news — suggests market has already substantially digested the "Anthropic filed first" signal. - Precedent: SpaceX IPO speculation running in parallel suggests markets are actively differentiating mega-cap tech IPO sequencing based on filing dates and CFO readiness signals, supporting weight on concrete filing/prep milestones over CEO rhetoric.

1. Both forecasters treat the 84% Kalshi price as strongly informative, but the brief notes OpenAI's confidential S-1 was actually filed earlier (~May 22 vs. Anthropic's June 1) and that "the market resolves on confirmed IPO, not filing" — this creates a structural mismatch between the narrative driving the price ("Anthropic filed first") and the actual resolution criterion, which neither forecast interrogates. 2. Neither forecast adequately weights the long horizon to 2040: given "no confirmed public S-1... actual listing dates unknown" and reports of OpenAI slipping "toward 2027," there is ample time for reversals, competitive catch-up, delisting risk, or even neither firm IPOing for years — this tail scenario is underweighted in both ~80-83% YES estimates. 3. The 90-day price range (22%–88%) reflects extreme volatility on sparse volume (~1,169 contracts/day), yet both forecasters treat the current 84% as a stable equilibrium rather than flagging the risk that sentiment could swing again on the next funding round, lawsuit development, or Altman announcement — this argues for wider uncertainty bands than either forecast implies. 4. Both forecasts cluster within 3 points of each other and almost exactly track the market price, suggesting herding on the Kalshi anchor rather than independent evidence synthesis; neither meaningfully engages with the countervailing case that OpenAI's stronger revenue momentum (July ARR "topped all of Q2") and Altman's explicit September target could flip sequencing. 5. The Pentagon legal dispute affecting Anthropic's CFO is noted as "unquantified" but material; both forecasts mention it only in passing without adjusting probability for the risk it materially delays Anthropic's readiness relative to OpenAI. 6. Neither forecast explores the scenario where both companies delay indefinitely or pursue alternative liquidity (large secondary sales, continued mega-rounds) instead of IPOing, which the brief's "No" outcome explicitly includes ("OpenAI IPOs first, or Anthropic does not IPO first") — this composite "No" case deserves more explicit decomposition than simply "OpenAI wins the race."

79% (medium confidence)

Yes: 79% No: 21%

The Kalshi anchor at 84% YES reflects a large, news-driven repricing following Anthropic's June 1, 2026 confidential S-1, its $965B round surpassing OpenAI's valuation, near-profitability, and the most concrete visible prep steps (ex-Airbnb IPO CFO, pre-IPO credit lines, investor meetings, reported October 2026 target). Both forecasters essentially adopted the market price; the critique correctly identifies that the resolution criterion is a confirmed IPO, not a filing date (where OpenAI actually filed ~May 22, earlier), and that 'No' is a composite bucket that also absorbs the tail where Anthropic never lists first (indefinite delay, alternative liquidity via mega-rounds/secondaries, acquisition, or Pentagon-litigation-driven slippage) — all of which favor slightly more mass on No than the raw anchor. Offsetting this, the horizon to 2040 is long enough that Anthropic's readiness lead is likely to convert, and OpenAI's reported slippage toward 2027 plus Friar's caution weakens the OpenAI-first case. I therefore settle modestly below both forecasts and the market at 0.79 YES, accepting a wider effective uncertainty band given thin volume (~1.2k contracts/day) and a 22%–88% 90-day range.

Total pipeline time: 257.1s