Scan: Will OpenAI release a social network in 2026?

1. Planner Fable 5.1 · Fermi decomposition + tool plan

Tools chosen: polymarket_direct claude_news article_search gdelt_news polymarket_related kalshi_related wikipedia

Sub-questions (Fermi decomposition)

What is the current Polymarket YES price and how has it moved recently?
Has OpenAI already released a product (e.g., the Sora app with its social feed, or a ChatGPT social feed) that OpenAI itself presents as a 'social network' or 'social platform,' and would it qualify under the resolution criteria?
What is the status of the reported OpenAI social network prototype (X-like feed with image generation) first reported in April 2025 — is it still in development, tested internally, or shelved?
Has Sam Altman or other OpenAI executives made public statements in the last 6-12 months about launching a social network or social features in 2026?
Have credible outlets (The Verge, The Information, Bloomberg) reported on internal timelines, hiring, or trademark filings related to an OpenAI social product?
Do related prediction markets (Kalshi, other Polymarket markets on OpenAI products) imply consistent or divergent probabilities?
How often have rumored 'social network' launches by big tech/AI companies actually shipped publicly within ~18 months of first reports (base rate)?

Planner reasoning

This is a Polymarket question about whether OpenAI launches a publicly accessible social network by end of 2026. Key uncertainties are whether reported prototypes (e.g., the X-like feed tied to ChatGPT image generation, the Sora app's feed) exist and how OpenAI frames them, plus whether the Sora app or similar already qualifies under the resolution rules. Start with the Polymarket price, then news on OpenAI social network plans, Sora app status, and cross-venue signals.

2. Research Parallel tool calls

tool	status	time	items	summary
polymarket_direct	OK	0.6s	1	## This Market's Polymarket Data Will OpenAI release a social network in 2026? - Current price (probability): 41.50% - 7-day price change: +36.00% - 30-day price change: +34.00% - Total volume: $70,881 (USD notional) - Price range: 5.50% - 41.50% - Data points: 90 days
claude_news	OK	34.4s	14	Findings: - April 2025 – Initial reports of X-like prototype: The Verge first reported OpenAI was building an X-like social network, with an internal prototype focused on ChatGPT's image generation that contains a social feed . It was unclear whether this would be a standalone app or inte
article_search	OK	10.5s	34	Fetched 34 unique articles across 4 queries (lookback=180d). 'OpenAI social network': 10 articles \| 'OpenAI Sora app social feed': 10 articles \| 'Sam Altman social network ChatGPT feed': 10 articles \| 'OpenAI X-like social media prototype': 10 articles
gdelt_news	OK	163.1s	10	GDELT: 10 articles across 3 queries (lookback=90d). 'OpenAI social network': error GDELT rate-limited after retries (429) \| 'OpenAI Sora app feed': error GDELT rate-limited after retries (429) \| 'OpenAI social platform launch': 10 hits
polymarket_related	OK	10.2s	1	Scanned 100 active Polymarket markets, kept 1 matches. keyword 'OpenAI': 1 markets \| keyword 'social network': 0 markets \| keyword 'Sora': 0 markets \| keyword 'Sam Altman': 0 markets
kalshi_related	OK	10.1s	2	2 related markets / summaries. keyword 'OpenAI': ok \| keyword 'social network': ok \| keyword 'Sora': no matches
wikipedia	OK	0.1s	3	Fetched 3 Wikipedia entries (0 missing pages).

3. Evidence Brief Sonnet 5 · 8414 chars

# Current state No Kalshi-direct price was returned for this specific ticker; the only quantitative anchor available is Polymarket, pricing YES at 41.5% (up sharply from a 5.5% low). OpenAI already shipped and branded one product ("Sora," Sept 2025) as a social app/network with feeds and profiles, but that app was discontinued by OpenAI in April 2026, and a separate "human-only" biometric social network remains in very early internal development as of Jan 2026 with no confirmed 2026 public launch. # Timeline of key events - 2025-02: Sam Altman jokes on X "ok fine maybe we'll do a social app" after Meta's assistant app launch (confirmed quote; freethink.com). - 2025-04-15: The Verge/TechCrunch report OpenAI is building an internal X-like prototype with an image-generation social feed; unclear if standalone or ChatGPT-integrated (reported). - 2025-04: Reports that Altman is privately soliciting outside feedback on a product that "could rival X" (reported; Slashdot). - 2025-09-30: OpenAI publicly launches "Sora," described by OpenAI itself as "a new social iOS app" with feeds, remixing, cameos; Bloomberg and Platformer both explicitly call it a social network/app (confirmed). - 2025-11-04: Sora comes to Android in select regions with no waitlist for US users (confirmed). - 2026-01-28: Forbes/Gizmodo report OpenAI is separately, quietly developing a "human-only" biometric social network (eye-scan/Face ID verification) via a team of <10 people; early stage, unclear ties to Sora/ChatGPT (reported). - 2026-03-26: CNN reports OpenAI is "throwing in the towel" on Sora amid cost pressures as it pivots to enterprise/compute priorities (reported). - 2026-04-03: CNN reports OpenAI "shut down its AI-generated video social media app Sora" amid refocus on compute-heavy core products (confirmed shutdown signal). - 2026-04-26: Sora web/app formally discontinued per OpenAI Help Center; API discontinuation set for 2026-09-24 (confirmed, Wikipedia/glbgpt.com). - 2026 (mid-year, ongoing): OpenAI news dominated by GPT-5.5, custom Jalapeño chip, IPO filing, enterprise pivot — no further reporting of new social-network launch (confirmed via article_search). # Event Will OpenAI launch a publicly-accessible social network (standalone or integrated) by December 31, 2026? # Outcomes to forecast - Yes - No # Kalshi market anchor No direct Kalshi price was retrieved for this ticker in the research (kalshi_direct tool output absent). Cross-market Kalshi search for "OpenAI"/"social network" returned no matching Kalshi market for this specific event — treat Polymarket as the best available anchor: **41.5% YES**, up from a 90-day low of 5.5%, +36% over 7 days and +34% over 30 days, on modest volume ($70,881 total). The sharp recent rally suggests repricing off new information (likely the Jan 2026 Forbes biometric-network report), not resolution of the underlying question. # Sub-question answers 1. **Polymarket YES price/trend?** — 41.5%, rallying hard (+36% 7d, +34% 30d) off a 5.5% low; thin volume ~$71k (polymarket_direct). 2. **Has OpenAI already released a product it calls a "social network"?** — Yes: Sora (launched Sept 30, 2025) was explicitly branded by OpenAI as "a new social iOS app" with feeds/remix/cameo features; Bloomberg and Platformer both called it a social app/network. This plausibly already satisfies the resolution language ("launches a social network by Dec 31, 2026"), though it predates 2026 and was later discontinued (April 2026), creating interpretive ambiguity (claude_news, Wikipedia). 3. **Status of the April 2025 X-like prototype?** — Unclear if it became Sora or is separate; no confirmed standalone launch of an "X-rival" as described in April 2025. A related but distinct "human-only" biometric social network was reported in Jan 2026, still in very early internal development (<10-person team) (claude_news, Forbes). 4. **Recent Altman/exec statements on social launch in 2026?** — Only Altman's Feb 2025 offhand joke found; no confirmed statements in the last 6-12 months committing to a 2026 social launch (claude_news). 5. **Reporting on internal timelines/hiring/trademarks?** — Forbes (Jan 2026) reports a small (<10 person) team building a biometric social platform; no hiring/trademark specifics or timeline found (claude_news). 6. **Related prediction markets consistent?** — Polymarket "Will OpenAI not IPO by Dec 31, 2026" prices IPO-non-completion at 94.3% (i.e., IPO seen as unlikely this year), consistent with OpenAI being IPO/enterprise-focused per CNBC reporting; Kalshi has no matching social-network market, only tangential OpenAI/Anthropic IPO-race and litigation markets (polymarket_related, kalshi_related). 7. **Base rate for rumored big-tech social launches shipping within ~18 months?** — Not directly answered by research; qualitatively, OpenAI's own Sora case shows a rumor (Apr 2025) converting to an actual branded social launch within ~5 months (Sept 2025), i.e., a positive precedent, though that product was later killed. # Key facts (high-confidence, factual) 1. [OpenAI/Bloomberg/Platformer] Sora launched Sept 30, 2025 as a self-described "social iOS app" with feeds/remix/cameo; expanded to Android Nov 2025 with no waitlist in the US. 2. [Wikipedia/CNN/glbgpt.com] Sora web/app was discontinued April 26, 2026; API discontinuation scheduled Sept 24, 2026, amid cost/compute reprioritization. 3. [Forbes/Gizmodo, Jan 2026] A separate, early-stage "human-only" biometric social network is being built by a small (<10) OpenAI team; no launch date given. 4. [CNBC/CNN, 2026] OpenAI's 2026 public narrative is dominated by enterprise pivot, GPT-5.5, custom chips, and IPO prep — not consumer social products. 5. [Polymarket_related] Market pricing implies OpenAI unlikely to IPO in 2026 (94% "not IPO"), reinforcing a company posture focused on infra/enterprise over new consumer social launches. # Cross-market signals - Kalshi related: No direct social-network market found; adjacent Kalshi markets (OpenAI/Anthropic IPO race, Apple v. OpenAI suit) show no social-product signal. - Polymarket: This market itself at 41.5% YES, sharply up recently; no other Polymarket market specifically tracks "social network" or "Sora." - Sportsbook implied: None available. # Analyst opinions and speculation - Platformer (Oct 2025) explicitly argued Sora already constitutes "a social network," a framing directly relevant to resolution. - Polymarket aggregator commentary (per claude_news) states "no official announcements... have signaled plans for a social network launch in 2026," viewing further launch as inconsistent with OpenAI's 2026 enterprise focus. - CNN (Mar/Apr 2026) frames Sora's discontinuation as OpenAI "throwing in the towel" on consumer social bets amid compute cost pressure — bearish for further social launches in 2026. # Directional lean per outcome - **Yes**: Sora already launched (Sept 2025) as an OpenAI-branded social app satisfying most definitional criteria, arguably already triggering resolution under "by Dec 31, 2026" language; separate biometric social project in development could also ship. Recent 30-point Polymarket rally suggests growing belief in eventual qualifying event. - **No**: Sora was discontinued in April 2026 and predates the 2026 calendar year of interest in colloquial framing; company strategy has visibly pivoted to enterprise/compute/IPO; biometric project is early-stage (<10 people) with no public timeline; market still prices YES well below 50%, reflecting real doubt Sora counts or that a new one ships by year-end. # Gaps / unknowns - No official Kalshi price for this exact ticker was retrieved — Polymarket used as proxy anchor. - Ambiguous whether Sora's 2025 launch (later discontinued) satisfies "launches... by Dec 31, 2026," which is the crux of this market's true probability. - No confirmation on biometric social network's 2026 launch timeline or public accessibility. - Unclear how resolution source (OpenAI official info + "consensus of credible reporting") will treat a discontinued but previously-launched social product. # Calibration anchors - Polymarket current YES price: 41.5% (anchor, no Kalshi-direct price available). - Precedent: OpenAI rumor-to-launch (Sora) took ~5 months (Apr 2025 report → Sept 2025 launch), then discontinued ~7 months later — showing both rapid shipping capability and rapid abandonment of consumer social bets.

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

claude-fable-5-1

40% conf

Yes	30%
No	70%

No Kalshi price exists; Polymarket prices YES at 41.5% after a sharp, thin-volume rally with no identified catalyst beyond the Jan 2026 Forbes report of a <10-person team on an early-stage biometric social network. The strongest 'Yes' case is definitional: Sora launched Sept 2025 as an OpenAI-branded social app, but it predates 2026 and was formally discontinued April 26, 2026, so under a natural reading of 'release in 2026' it likely does not count. OpenAI's 2026 posture (enterprise pivot, GPT-5.5, chips, IPO prep, CNN's 'throwing in the towel' on consumer social) makes a new public social-network launch by Dec 31 plausible but unlikely, though OpenAI showed it can ship a social product within ~5 months of a rumor. I shade below the Polymarket anchor to ~30% to reflect the resolution ambiguity around Sora while weighting the bearish strategic signals.

gpt-5.6-sol

67% conf

Yes	31%
No	69%

The best available market anchor is Polymarket at 41.5% YES, though its modest volume and sharp recent rally make it a noisy substitute for a missing Kalshi price. Sora demonstrates OpenAI can quickly turn a prototype into a social product, but it launched in 2025 and was discontinued in April 2026, so it likely does not satisfy a release specifically “in 2026.” The separate biometric network remained an early-stage project with fewer than 10 people and no public timeline, while OpenAI’s enterprise and compute pivot points toward No. With only a few months remaining and no subsequent launch reporting, I discount YES below the market anchor while retaining material probability for a rapid or integrated release.

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

Both forecasters converge tightly (0.30/0.31 YES) despite acknowledging significant resolution ambiguity around Sora—neither explores the scenario where resolution criteria explicitly count "launched by Dec 31 2026" (i.e., cumulative, not "active as of Dec 31"), which per the brief's own event framing ("launch a publicly-accessible social network... by December 31, 2026") would seem to favor YES since Sora unambiguously launched and was publicly accessible within the window (Sept 2025–Apr 2026), even though later discontinued; this reading is arguably more natural than the "must still exist in 2026" interpretation both forecasters lean on.

Neither forecaster addresses the event's explicit "standalone or integrated" clause from the Event definition, which could make ChatGPT-integrated social features (feeds, sharing, profiles) sufficient for YES even without a dedicated app — a lower bar than either analysis considers.

Both forecasters shade below the Polymarket anchor for similar reasons (resolution ambiguity, enterprise pivot) without flagging that the anchor itself is thin/noisy ($71k volume, +34% 30-day rally) and may not reflect informed money — yet both still treat 41.5% as a meaningful reference point to discount from, which is somewhat inconsistent with dismissing its reliability.

Neither forecast weighs the biometric social network's optionality with enough nuance: given OpenAI's demonstrated ~5-month rumor-to-launch speed (Sora precedent) and Jan 2026 reporting confirming active development, there's roughly 11 months left in 2026 — enough time for a rushed launch — but both treat this scenario as low-probability without quantifying it explicitly.

Both rationales converge on nearly identical language and numbers, suggesting anchoring on each other's reasoning rather than independent triangulation; neither offers a differentiated tail scenario (e.g., OpenAI re-launching a slimmed-down Sora, or a competitor-forced pivot back to consumer social) that could push probability meaningfully higher or lower.

Neither forecast explicitly grapples with the "consensus of credible reporting" resolution source mentioned in the Gaps section, which could resolve the Sora-counts-or-not ambiguity in either direction — this is a critical unresolved crux that deserves explicit probability-weighting rather than a single-point "likely doesn't count" assumption.

6. Aggregator Fable 5.1 · final calibrated forecast

34% (54% confidence)

Yes: 34% No: 66%

The event title specifies a release 'in 2026', and the Polymarket contract traded at a 5.5% low while Sora was live and publicly accessible, which strongly implies traders and the resolution rules do not count Sora's Sept 2025 launch — so the critique's 'Sora already satisfies it' argument is mostly neutralized by revealed market pricing. Sora's April 2026 discontinuation and OpenAI's enterprise/compute/IPO pivot are bearish for a new consumer social launch, but the sharp thin-volume rally from 5.5% to 41.5% has no identified catalyst in the brief, which raises the possibility of unreported new information (e.g., progress on the biometric 'human-only' network or ChatGPT-integrated social features), and OpenAI has shown it can ship a social product within ~5 months of a rumor. The 'standalone or integrated' clause also lowers the bar somewhat. I therefore stay a bit above both forecasters (who may be over-discounting a market that is rallying on unknown news) but below the noisy Polymarket anchor, at ~34% YES.

Pipeline Timing

Total pipeline time: 275.6s

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