In late August 2026 OpenAI announced on its own site and X (reported by Bloomberg) that it will wind down its contract supplying models to Cursor — Anysphere's AI coding tool — with a proposed shutoff date of Nov 12, 2026, following SpaceX's ~$60B acquisition of the company that closed earlier that month. OpenAI says the custom agreement gives it only a limited cancellation window after a change of control and that it 'cannot be confident that SpaceX will use our technology within our terms of service' — CNBC quotes it citing Elon Musk's companies' history of contract violations — and it will withhold future models, including Astra, from Cursor. The immediate operational hit looks bounded: Cursor CEO Michael Truell says OpenAI models serve only ~5% of user traffic (denominator unverified), and Dealroom notes Cursor also resold Anthropic's Claude, though OpenAI has not detailed which capabilities vanish Nov 12 and these snippets say nothing about Cursor's replacement routing (xAI Grok, Anthropic capacity) or actual user impact. The case file adds a Reuters-reported Anthropic compute expansion and a single unreproduced BYOK incompatibility report as the open impact questions, with the Nov 12 shutoff itself the next meaningful checkpoint.
2026-10-05T18:30:58Z
grounded: converges/high — Converges: the Nov 12 shutoff is Scott's vendor-disappearance/exit test run in public — Model Perishability and Sovereign Software Assurance made concrete — and
2026-10-05T18:21:39Z
First direct user-experience data on Cursor's substitute path: an Ask HN report finds Cursor's native xAI models competitively usable ahead of the Nov 12 shutoff — the first weak signal that replacement capacity is materializing, exactly the trigger this case pre-committed to as material, though single-source and no evidence of parity. The case stays a dated implementation watch; the multi-platform spread reading reflects the stale late-August announcement burst, and with current rates near zero and the periphery no longer expanding, low heat stands.
2026-10-05T17:32:20Z
evidence attached: hn.story.49966668 — First-hand user report that Cursor's native xAI models deliver competitive usage and agentic experience — direct evidence on Cursor's substitute model access and competitive position after the OpenAI split.
2026-09-11T10:28:45Z
The staleness check adds no substantive evidence, leaving this a pending provider-access transition rather than an observed disruption. The November 12 cutoff still warrants implementation monitoring, but Cursor’s replacement capacity and the reproducibility of the isolated BYOK failure remain unresolved.
2026-09-09T10:23:57Z
Refreshed Wired discussion adds financial and strategic speculation, not evidence of replacement capacity, migration, or changed access. The planned cutoff remains established, but neither the ambiguous traffic-share claim nor the isolated BYOK report establishes the operational impact; keep this on a slow implementation watch.
2026-09-07T09:28:06Z
The staleness check brings no new evidence: the planned cutoff remains established, while the single-user BYOK incompatibility report still does not demonstrate cutoff-driven access loss. Keep this as a pending implementation story rather than a developing disruption; replacement capacity, reproducible portability tests, or changes to the November 12 deadline would materially change its meaning.
2026-09-05T08:26:17Z
The refreshed comments neither reproduce nor technically challenge the reported BYOK failure, leaving it a provisional portability warning rather than evidence of cutoff-driven access loss. The planned termination remains established, but no new migration or user-impact facts have emerged.
2026-09-05T05:23:14Z
A single-user diagnostic report introduces a concrete possible limit to BYOK as an exit path: API-protocol and model-registry compatibility may block access even when credentials work independently. This makes portability testing more actionable for Scott, but the truncated account does not establish a reproducible workaround, broad access loss, or a causal link to the partnership termination.
2026-09-05T05:22:04Z
evidence attached: reddit.post.1w7rkla — A concrete BYOK workaround exposes a Responses-API and model-registry incompatibility that materially contextualizes Cursor's changing access to OpenAI models.
2026-09-04T16:35:07Z
The Reddit velocity spike is renewed amplification of Wired’s coverage, not a new operational development. The cutoff remains established, while Cursor’s replacement routing, capacity, and user impact remain unresolved pending direct evidence or implementation.
2026-09-04T11:29:44Z
The refreshed discussion adds no operational facts beyond the established November 12 cutoff and speculative financial or strategic interpretations. The case remains a concrete provider-portability example, but its practical significance depends on direct evidence of Cursor’s replacement routing, capacity, or user impact.
2026-09-04T06:28:05Z
The latest Hacker News item only redistributes Wired’s coverage of the established November 12 cutoff; it adds no operational evidence about Cursor’s replacement capacity, routing, or user impact. The case remains dormant pending a direct migration update or implementation effects.
2026-09-04T06:21:51Z
evidence attached: hn.story.49561018 — shared external link with case evidence
2026-09-04T00:29:13Z
The new Hacker News item merely redistributes Wired’s follow-on coverage of the established November 12 cutoff and adds no evidence about replacement routing, capacity, or user impact. The case remains dormant pending a direct operational update or implementation effects.
2026-09-03T19:23:26Z
evidence attached: hn.story.49555073 — shared external link with case evidence
2026-09-03T17:57:57Z
Wired’s follow-on coverage reinforces the established provider cutoff but adds no new facts about Cursor’s replacement capacity, routing, or user impact. The case remains a concrete provider-portability warning rather than an actively developing operational disruption.
2026-09-03T17:24:35Z
evidence attached: reddit.post.1w6cqrr — Independent coverage materially corroborates the reported OpenAI-Cursor partnership termination and its developer-tool implications.
2026-09-01T17:44:19Z
The newly attached Hacker News item is duplicate coverage of the established November 12 cutoff, not fresh evidence about Cursor’s replacement capacity, routing, or user impact. The case remains a provider-portability warning but is operationally dormant until a direct migration update or implementation effects emerge.
2026-09-01T17:26:48Z
evidence attached: hn.story.49524784 — This is independent corroboration of the reported OpenAI-Cursor model-access cutoff and its potential competitive impact on AI developer tools.
2026-09-01T13:38:41Z
The new Hacker News item is another pointer to the already-established November 12 cutoff and adds no operational evidence about Cursor’s replacement routing, capacity, or user impact. The case remains dormant pending a direct migration update or implementation effects.
2026-09-01T13:24:51Z
evidence attached: hn.story.49521111 — This independently corroborates the reported end of OpenAI model access for Cursor, although the linked source provides limited detail.
2026-09-01T01:27:59Z
The new Hacker News item is another pointer to the already-established OpenAI cutoff and adds no migration, routing, or user-impact evidence. The case remains dormant until Cursor or Anthropic details replacement capacity or the November 12 cutoff produces operational effects.
2026-09-01T01:22:36Z
evidence attached: hn.story.49516583 — shared external link with case evidence
2026-08-31T16:39:10Z
The attached Hacker News item is another pointer to Reuters’ already-captured corroboration, not a new operational development. The cutoff is established, but Cursor’s replacement routing and user impact remain unresolved until a direct migration update or implementation evidence emerges.
2026-08-31T16:24:40Z
evidence attached: hn.story.49511084 — Reuters independently corroborates the reported end of OpenAI's Cursor partnership and its implications for AI developer-tool model access.
2026-08-31T05:22:54Z
The Reddit velocity spike is renewed amplification of the established cutoff, not evidence of Cursor’s migration plan, model-routing changes, or user impact. The case remains dormant pending a direct operational update or implementation near the November deadline.
2026-08-30T23:32:20Z
The additional Hacker News item only redistributes OpenAI’s already-established, dated contract wind-down and adds no migration, routing, or user-impact evidence. The case remains a concrete provider-portability warning but is operationally dormant pending a direct Cursor response or implementation change.
2026-08-30T23:23:23Z
evidence attached: hn.story.49503609 — Independent Hacker News coverage of OpenAI’s announcement corroborates the open case and its implications for Cursor’s model access.
2026-08-30T21:30:45Z
The refreshed comments add only repetitive speculation and no evidence about Cursor’s replacement routing, the meaning of its reported 5% OpenAI traffic share, or user impact. The termination remains established, but the operational story is dormant pending a direct migration plan or cutoff implementation.
2026-08-30T04:26:35Z
Refreshed engagement and comments add only repetitive interpretation of the established cutoff and ambiguous 5% traffic claim. Cursor’s replacement routing and actual user impact remain unresolved, but there is no fresh operational development.
2026-08-30T00:28:23Z
Refreshed comments continue to reinterpret the ambiguous 5% traffic figure and speculate about motives, without adding evidence on replacement routing or user impact. The cutoff remains established, but the operational story is dormant pending a direct Cursor or Anthropic plan or implementation evidence.
2026-08-29T23:24:58Z
The new items firm up the contractual mechanism and November 12 cutoff while adding a Reuters-attributed indication that Anthropic may expand compute support for Cursor. This points toward managed provider substitution rather than acute disruption, but direct migration details and observed user impact remain absent.
2026-08-29T21:23:34Z
evidence attached: hn.story.49493319 — This independently corroborates the reported loss of OpenAI model access and its implications for Cursor’s competitive position.
2026-08-29T20:23:31Z
evidence attached: reddit.post.1w1w7f9 — This independently corroborates the reported shutdown date and adds concrete contract and model-portability implications for Cursor.
2026-08-29T17:29:55Z
The refreshed discussion remains repetitive amplification of the established termination and ambiguous 5% traffic claim, with no direct evidence of Cursor’s replacement routing or user impact. The case remains dormant pending an operational response or cutoff implementation.
2026-08-29T14:26:34Z
Refreshed comments continue to debate motives and the ambiguous 5% traffic figure without adding direct evidence on Cursor’s routing plan or user impact. The termination remains established, but the case is dormant until an operational response or cutoff implementation emerges.
2026-08-29T13:26:35Z
Refreshed comments and engagement remain repetitive amplification, adding no verified detail about Cursor’s exposure, replacement routing, or user impact. The termination is established, but the case remains dormant pending a direct Cursor response or operational evidence.
2026-08-29T12:29:36Z
Refreshed comments and minor engagement add only speculation about motives and the already-reported 5% traffic share. The cutoff remains established, but Cursor’s actual exposure, migration plan, and user impact remain unresolved pending direct evidence or implementation.
2026-08-29T11:34:42Z
The newly attached post repeats the second-hand claim that OpenAI models represent about 5% of Cursor traffic but adds no verification, denominator, or operational detail. The cutoff remains established, while the apparent evidence of provider resilience is still provisional and the case remains dormant pending Cursor’s response or implementation.
2026-08-29T11:23:23Z
evidence attached: reddit.post.1w1jihk — This provides independent user-facing corroboration and a potentially important estimate of how little Cursor depended on OpenAI traffic.
2026-08-29T10:31:47Z
Refreshed comments and engagement only amplify established facts and speculation; they add no verified migration plan, traffic-share evidence, or observed user impact. The case remains dormant until Cursor responds or the November cutoff approaches.
2026-08-29T09:26:23Z
The newly attached coverage and refreshed discussion merely repeat the established cutoff, adding no verified detail on Cursor’s exposure, replacement routing, or user impact. The case remains a concrete provider-dependency example but is dormant until Cursor responds or implementation begins.
2026-08-29T09:23:21Z
evidence attached: reddit.post.1w1hk4c — shared external link with case evidence
2026-08-29T08:31:06Z
Refreshed discussion remains speculative and adds no evidence about Cursor’s migration plan, actual OpenAI traffic share, or user impact. The cutoff is established, but this episode is now waiting on implementation rather than actively developing.
2026-08-29T07:29:10Z
Refreshed comments add speculation about motives and user effects but no new facts about the cutoff, Cursor’s exposure, or migration plans. The termination remains established while its practical impact is still unresolved.
2026-08-29T06:31:23Z
Reuters independently confirms the planned OpenAI–Cursor termination, making the provider cutoff established rather than dependent on reconstructed testimony or social amplification. Attention now shifts to implementation and user impact, which remain uncertain given the unverified claim that OpenAI models account for only about 5% of Cursor traffic.
2026-08-29T06:23:16Z
evidence attached: hn.story.49487134 — Independent Reuters corroboration of the reported OpenAI–Cursor partnership termination, materially strengthening the open case.
2026-08-29T05:30:08Z
The new Reddit post and refreshed comments add amplification rather than independent facts, leaving the cutoff established but its practical impact still poorly quantified. With no implementation details or first-party confirmation of the reported 5% traffic share, the case cools while awaiting evidence of migration or user impact.
2026-08-29T05:23:27Z
evidence attached: reddit.post.1w1dbxv — The post appears to provide user-level corroboration of the reported OpenAI-Cursor deal ending, though with limited detail.
2026-08-29T04:29:51Z
The reported 5% traffic share suggests OpenAI’s cutoff may reduce model choice more than disrupt Cursor’s operations, shifting the case from immediate platform vulnerability toward evidence of successful provider diversification. The figure remains second-hand and lacks timeframe or denominator, so the magnitude of that resilience is not yet established.
2026-08-29T04:23:20Z
evidence attached: reddit.post.1w1bly3 — Cursor’s CEO reportedly saying OpenAI models serve only about 5% of traffic materially contextualizes the competitive impact of the partnership ending.
2026-08-29T03:30:26Z
Bloomberg’s reporting and reconstructed testimony of OpenAI’s own announcement provide two distinct lines supporting a planned model-supply wind-down, moving this beyond a headline-only rumor. The cutoff’s implementation and impact on Cursor remain unsettled, but the provider-dependency risk is now concrete.
2026-08-29T03:29:00Z
grounded: converges/medium — The reported cutoff is a concrete, consequential instance of Scott’s Model Perishability and Sovereign Software Assurance claims: coding tools need swappable pr
2026-08-29T03:27:36Z
origin walked (codex/luna, conf 0.99): anchor hn.story.49486444 -> echo.blog.3ac5377561 by OpenAI
2026-08-29T03:26:25Z
case created — The reported termination is a consequential but currently single-source change in a major coding-tool platform relationship.