Scan: OpenAI announces it has achieved AGI before 2027?

date	predicted	market price	confidence
2026-09-05	21%	26%	54%
2026-08-26	7%	10%	73%

1. Planner Fable 5.1 · Fermi decomposition + tool plan

Tools chosen: polymarket_direct polymarket_related kalshi_related article_search claude_news gdelt_news wikipedia

Sub-questions (Fermi decomposition)

What is the current Polymarket price and trend for this market, and what do related AGI/OpenAI markets on Polymarket and Kalshi imply?
What have Sam Altman and other OpenAI officials said in 2025-2026 about AGI timelines, and have they moved toward or away from declaring AGI achieved?
Under the October 2025 OpenAI-Microsoft restructuring, does an AGI declaration require verification by an independent expert panel, and how does that constrain OpenAI's incentive/ability to announce AGI before 2027?
Has any OpenAI representative made statements (e.g., 'we've basically achieved AGI', 'AGI is here') that credible reporting has interpreted as an AGI announcement, and how did markets react?
What are OpenAI's next major model releases (GPT-6 or similar) expected before end of 2026, and does OpenAI frame them with AGI language?
What are the financial/legal incentives for OpenAI to declare AGI (Microsoft IP rights cutoff, revenue share) versus to avoid declaring it?

Planner reasoning

This is a Polymarket question about whether OpenAI officially announces AGI before end of 2026. Key drivers are OpenAI leadership statements on AGI timelines, the Microsoft contract's AGI clause (and the expert panel introduced in the Oct 2025 restructuring), and how liberally 'announces' would be interpreted. The market price is the primary anchor; cross-market checks on Kalshi/Polymarket AGI questions and recent news on Altman's rhetoric are the main empirical inputs.

2. Research Parallel tool calls

tool	status	time	items	summary
polymarket_direct	OK	0.8s	1	## This Market's Polymarket Data OpenAI announces it has achieved AGI before 2027? - Current price (probability): 17.50% - 7-day price change: -10.00% - 30-day price change: +9.00% - Total volume: $256,498 (USD notional) - Price range: 7.00% - 27.50% - Data points: 89 days
polymarket_related	OK	3.8s	1	Scanned 100 active Polymarket markets, kept 1 matches. keyword 'AGI': 1 markets \| keyword 'OpenAI': 0 markets \| keyword 'Sam Altman': 0 markets \| keyword 'superintelligence': 0 markets
kalshi_related	OK	3.6s	3	3 related markets / summaries. keyword 'AGI': ok \| keyword 'OpenAI': ok \| keyword 'artificial general intelligence': ok
article_search	OK	0.6s	29	Fetched 29 unique articles across 4 queries (lookback=60d). 'OpenAI announces AGI': 10 articles \| 'Sam Altman AGI timeline 2026': 10 articles \| 'OpenAI Microsoft AGI clause expert panel': 10 articles \| 'OpenAI AGI declaration': 10 articles
claude_news	OK	28.6s	15	Here are the key findings for the OpenAI AGI forecasting question: Altman/OpenAI statements on AGI status (2025-2026): - In his January 2025 "Reflections" blog post, Altman wrote: "We are now confident we know how to build AGI as we have traditionally understood it." He added that OpenAI is
gdelt_news	OK	102.9s	20	GDELT: 20 articles across 2 queries (lookback=30d). 'OpenAI AGI achieved': 10 hits \| 'Altman AGI announcement': 10 hits
wikipedia	OK	0.2s	3	Fetched 3 Wikipedia entries (0 missing pages).

3. Evidence Brief Sonnet 5 · 8142 chars

# Current state No AGI announcement has been made by OpenAI or an authorized representative. OpenAI executives (Brockman, Altman) have made informal, hedged statements suggesting capability milestones ("AGI era," Astra "might be AGI") but these are not formal declarations, and the contractual AGI-verification mechanism with Microsoft has been effectively defanged. Polymarket prices this event at 17.5%, down sharply from recent highs. # Timeline of key events - 2025-01: Altman writes "we are now confident we know how to build AGI as we have traditionally understood it" but stops short of declaring it achieved (confirmed, samaltman.com). - 2025-10-28: Microsoft-OpenAI restructuring deal states any future AGI declaration by OpenAI must be verified by an independent expert panel before triggering contractual consequences (confirmed, Microsoft blog). - 2026-02-27: Microsoft/OpenAI jointly reaffirm "AGI definition and processes are unchanged" (confirmed). - 2026-04-27: Microsoft and OpenAI amend the deal again — Microsoft's IP license extended to 2032 (non-exclusive), revenue share continues through 2030 regardless of AGI status, capped; commentators say this severs the commercial trigger tied to an AGI declaration ("AGI clause is dead") (confirmed/reported, Simon Willison, dev.to). - 2026-08-19: OpenAI CFO says company expects IPO in 2027 or sooner; no AGI claim (confirmed, CNBC). - 2026-08-23/24: Altman publicly admits he was "very wrong" on some AI timeline/adoption predictions; timeline lengthened rhetoric (reported, AEI/ToI). - 2026-09-03/04: OpenAI rolls out GPT-6 "Astra"; Brockman calls it a "turning point," says "we are now in the AGI era"; separately reported he thinks Astra "might be AGI" (reported, CNN/Fool.com) — not an official corporate AGI declaration. - 2026-08-29 to 09-07: Nvidia's Jensen Huang claims "AGI has already arrived" / "AI has achieved AGI in some tasks" — third-party claim, not from OpenAI (reported, multiple outlets). - 2026-09-10: CNN explainer notes no formal, testable AGI definition exists industry-wide, underscoring ambiguity of any future "announcement." # Event Will OpenAI or an official representative formally announce it has achieved AGI before Dec 31, 2026 (ET)? # Outcomes to forecast - Yes - No # Kalshi market anchor This ticker is tracked here via Polymarket data (no distinct Kalshi order book found for this exact ticker); Polymarket price is 17.5% YES, down 10pts over 7 days but up 9pts over 30 days, range 7%-27.5% over 89 days, $256k volume — treat as the consensus benchmark. A directly related Kalshi market ("When will any company achieve AGI — before Oct 1, 2026") prices only 4% YES, down 3pts (7d) and 2pts (30d), suggesting a more skeptical Kalshi-side consensus on any near-term AGI achievement claim, though that market has an earlier close date. # Sub-question answers 1. **Polymarket/related markets** — Polymarket prices this event at 17.5% (down from ~27.5% high, up from 7% low); no other Polymarket AGI/OpenAI markets exist. Kalshi's "any company achieves AGI before Oct 2026" trades at just 4%, implying a much lower probability for a verified/consensus AGI claim than this looser "announcement" market. [polymarket_direct, kalshi_related] 2. **Altman/OpenAI 2025-2026 statements** — Altman moved from confident ("we know how to build AGI," Jan 2025) to more hedged/skeptical by mid-2026, admitting he'd been "very wrong" on timelines and calling AGI "a sloppy term." Brockman, not Altman, made the most AGI-adjacent claim ("we are now in the AGI era," "Astra might be AGI") around the Sept 2026 GPT-6 Astra launch — sentiment is mixed, trending toward hedging rather than firm declaration. [claude_news, gdelt_news] 3. **Microsoft restructuring/expert panel** — Yes: the Oct 2025 deal requires an independent expert panel to verify any OpenAI AGI declaration before it has contractual effect. But by April 2026 Microsoft and OpenAI amended the deal to decouple IP rights/revenue share from the AGI trigger entirely, removing the main commercial incentive/consequence structure tied to a declaration. [claude_news/Microsoft blog, Simon Willison] 4. **Statements interpreted as announcements** — Brockman's "AGI era" and "Astra might be AGI" comments (Sept 2026) generated market buzz but were not treated by credible reporting as an official corporate declaration; Nvidia's Huang made separate, non-OpenAI claims that AGI "has arrived." No confirmed formal OpenAI announcement exists. [gdelt_news] 5. **Upcoming model releases/AGI framing** — GPT-6 "Astra" already launched (Sept 2026), OpenAI's most advanced model to date, framed with AGI-adjacent language by Brockman but not officially labeled AGI by OpenAI corporately; GPT-5.6 series preceded it (Jul 2026) with cost-cutting framing, no AGI claims. [article_search] 6. **Financial/legal incentives** — Post-April 2026 amendment, the AGI declaration no longer triggers loss of Microsoft's IP access or ends revenue share (both now run through 2030-2032 regardless), significantly reducing OpenAI's incentive to formally declare AGI; the expert-panel verification requirement also adds friction/delay against a unilateral declaration. Net effect: reduced incentive and added procedural hurdle work against a Yes resolution before 2027. [claude_news] # Key facts (high-confidence, factual) 1. [polymarket_direct] Polymarket YES price: 17.5%, 7d -10pt, 30d +9pt, volume $256k. 2. [kalshi_related] Kalshi "any company achieves AGI before Oct 2026" trades at 4% YES. 3. [Microsoft blog, Oct 2025] AGI declarations require independent expert panel verification under restructured Microsoft deal. 4. [Simon Willison, Apr 2026] Microsoft-OpenAI deal amended to decouple IP/revenue terms from AGI declaration. 5. [CNBC, Sept 2026] GPT-6 "Astra" launched; Brockman called it "AGI era" but no formal corporate AGI declaration reported. # Cross-market signals - Kalshi related: "Any company achieves AGI before Oct 1 2026" at only 4% — much lower than this event's implied 17.5%, suggesting Kalshi crowd is more skeptical of near-term genuine AGI achievement than Polymarket is of OpenAI merely "announcing" it. - Polymarket: This market itself (17.5%) shows volatility (7-27.5% range), recent 7-day decline suggests fading enthusiasm despite GPT-6 Astra hype. - Sportsbook implied: None available. # Analyst opinions and speculation - 80,000 Hours (Aug 2026): insider sentiment shifted toward longer AGI timelines after reasoning models plateaued on generalization tasks. - TechRadar/dev.to: analysts view the expert-panel and later de-linking of AGI clause as OpenAI/Microsoft mutually defusing the "AGI trigger," reducing pressure for a formal announcement. - Fool.com: framed Brockman's Astra comments as market-moving AI-trade catalyst, but stopped short of calling it an official AGI declaration. # Directional lean per outcome - **Yes**: Supported by escalating AGI-adjacent rhetoric (Brockman "AGI era," Huang "AGI arrived"), rapid capability releases (Astra/GPT-6), and OpenAI's competitive/marketing incentive to claim leadership. - **No**: Supported by removal of financial/contractual triggers for declaration (Apr 2026 amendment), expert-panel procedural hurdle, Altman's own increasing hedging/timeline lengthening, lack of any formal announcement through Sept 2026, and low correlated Kalshi price (4%) on genuine AGI achievement. # Gaps / unknowns - No confirmed Kalshi native order book price for this exact ticker (relied on Polymarket as proxy per data). - Unclear whether an "unofficial" strong claim (e.g., Brockman's) would satisfy resolution criteria requiring "OpenAI or official representative" announcement — ambiguity in interpretation could swing resolution. - No visibility into planned GPT-7 or further 2026 Q4 releases beyond Astra. # Calibration anchors - Polymarket YES price (anchor): 17.5%, recent range 7-27.5%. - Kalshi comparable market ("any company AGI before Oct 2026"): 4%. - Precedent: no AI lab has ever formally, verifiably declared AGI achieved; informal claims (Huang, Brockman) have not been treated as resolving events in past similar markets.

4. Forecasters Fable 5.1 + GPT-5.6 Sol · in parallel

claude-fable-5-1

58% conf

Yes	12%
No	88%

Polymarket anchors this at 17.5% YES, but the brief surfaces structural reasons to shade lower: the April 2026 Microsoft amendment decoupled IP/revenue from any AGI declaration, removing the main commercial incentive, and the expert-panel verification requirement adds procedural friction against a unilateral formal declaration. Altman has trended toward hedging (calling AGI a 'sloppy term,' admitting he was 'very wrong' on timelines), and the GPT-6 'Astra' launch—OpenAI's biggest 2026 release—produced only informal Brockman remarks ('AGI era,' 'might be AGI') that credible reporting did not treat as a corporate declaration. The related Kalshi market on any company achieving AGI before Oct 2026 sits at 4%, and no lab has ever formally declared AGI. Residual YES probability reflects resolution ambiguity around escalating executive rhetoric and possible Q4 releases, so I land at ~12%, modestly below the market.

gpt-5.6-sol

72% conf

Yes	14%
No	86%

The 17.5% Polymarket YES price is the primary consensus anchor, while Kalshi's related 4% market indicates greater skepticism toward a genuine near-term AGI claim. OpenAI's strongest recent language—Brockman's "AGI era" and "might be AGI" remarks around Astra—remains hedged and was not reported as a formal declaration, while Altman's rhetoric has become more cautious. The Microsoft amendments removed much of the commercial incentive to declare AGI, and expert-panel verification adds procedural friction. With Astra already launched, no formal announcement by mid-September, and only the remainder of 2026 available, I move modestly below the market anchor, while retaining meaningful YES probability because the undefined resolution standard could admit a strong executive statement.

5. Devil's Advocate Sonnet 5 · red-team critique

Both forecasters converge tightly (12% vs 14%) without meaningfully engaging with the large gap between Polymarket (17.5%) and the more skeptical Kalshi-analog (4%) — neither explains why they land closer to Polymarket than to the arguably more relevant "any company achieves AGI" benchmark, despite citing it prominently as evidence of stricter consensus.

Neither forecast grapples with the resolution-ambiguity issue in proportion to its weight in the brief: the brief explicitly flags that Brockman's "AGI era"/"might be AGI" comments could arguably satisfy a loose reading of "OpenAI or official representative announces," yet both treat this as only a minor residual factor rather than a genuine bimodal risk (resolves YES on lenient interpretation vs NO on strict interpretation).

Both forecasts underweight the base-rate/reference-class point buried in the brief — "no AI lab has ever formally, verifiably declared AGI achieved" — which is arguably the single strongest structural prior, yet it's mentioned only as one bullet among many rather than anchoring the estimate near single digits.

Neither forecaster explores the tail scenario of a strategic/marketing-driven announcement independent of contractual incentives — e.g., competitive pressure from Gemini/Anthropic or IPO narrative-building (CFO mentioned 2027 IPO plans) could motivate a splashy AGI claim in Q4 2026 purely for hype, which the removal of Microsoft's trigger actually makes *easier* (no verification consequence to fear) rather than harder.

Both treat the Oct 2025 expert-panel requirement as a hurdle against declaration, but neither notes the brief's ambiguity about whether the panel gates the *announcement itself* versus merely its *contractual consequences* — a formal OpenAI announcement could still happen and trigger this market's YES even if the panel later disputes it, which would make the procedural-friction argument weaker than both forecasts imply.

Neither forecast addresses the 7-day Polymarket price drop (27.5%→17.5%, -10pts) as a potentially overreacting signal versus the 30-day uptrend (+9pts) — this volatility (7%-27.5% range) suggests real uncertainty about interpretation that warrants wider confidence intervals than the narrow 12-14% band both forecasters converge on.

6. Aggregator Fable 5.1 · final calibrated forecast

14% (65% confidence)

Yes: 14% No: 86%

Both forecasters landed modestly below the 17.5% Polymarket anchor and the critique's points largely offset: the base-rate/Kalshi-skepticism arguments push lower, while the observations that the expert panel gates only contractual consequences (not the announcement itself) and that the April 2026 decoupling removes downside from a hype-driven declaration push higher. Polymarket traders have already observed Brockman's 'AGI era'/'might be AGI' remarks and still price only 17.5%, indicating those statements have not been treated as resolving and the remaining ~3.5 months carry limited but real tail risk of a marketing-driven or IPO-narrative announcement. Altman's increasing hedging, the absence of any formal declaration through the flagship GPT-6 launch, and the fact that no lab has ever formally declared AGI keep the structural prior firmly on No. I settle at 14% Yes, slightly below market to reflect the skeptical Kalshi cross-signal while preserving mass for resolution ambiguity and Q4 surprises.

Pipeline Timing

Total pipeline time: 193.9s

Per-tool research timings shown in the Research section above.