AI incidents: ChatGPT, Runway and Compliance API recoveries — 10 October 2026

Publisher disclosure: MUVO offers competing music, remix, and video tools. This report summarizes public sources and provider documentation; it is not an independent product test or a claim that MUVO is free of similar issues. Read our reporting method.

Incident status at a glance

Update, 13 October: OpenAI marked the Compliance API delay resolved on 12 October at 19:02 Istanbul. The original edition recorded three recovered events and that reporting delay under monitoring. This edition covers evidence checked from 7 October at 11:58 to 10 October at 11:49 Europe/Istanbul (UTC+3). The overlap catches missed events; it does not mean every disruption began today.

These are notice timestamps, not measured outage durations. Actual customer-impact start and end times are not established by these notices. Recovery claims come from the providers and do not independently confirm every account’s experience. Product launches belong in our separate 10 October AI news report.

ChatGPT Android: unavailable content recovered after an app fix

OpenAI marked the Android content issue resolved at 03:26 Istanbul on 10 October. Its initial notice, at 00:00 Istanbul, described some content being replaced by a message saying it was unavailable in that app version. The affected component was ChatGPT Conversations. Original incident and updates.

OpenAI said a fix was rolling out in the latest Android version and offered ChatGPT on the web as a temporary alternative for users of older versions. It did not name a specific app build, region or root cause. The final notice says affected services recovered; it does not establish that every Android user was affected or that the whole ChatGPT service was offline.

If the symptom persists, check your Android app version against the available update and use the provider’s web alternative while consulting the incident notice. This report does not establish lost conversation data.

ChatGPT plugins: workspace administration restored

OpenAI resolved a separate problem with viewing and managing workspace plugin availability and settings. The first acknowledgement was 00:48 Istanbul on 10 October; the provider reported recovery under monitoring at 01:52 and resolution at 02:26. Original plugin administration incident.

Some workspace administrators could not manage these settings, and some users could see plugins as unavailable. OpenAI listed Connectors/Apps as the affected ChatGPT component. It did not publish a cause, a region list or a specific workaround. The notice supports a scoped workspace/plugin problem, not a claim that all ChatGPT conversations failed.

Runway: app and API event resolved, with limited public detail

Runway acknowledged a problem at 01:44 Istanbul and marked it resolved at 02:10 on 10 October. Its official archive records partial-outage transitions for the App, Backend and Public API components, followed by their return to operational status. Incident notice and official incident archive.

The notice only says users were affected; it does not identify failing operations, particular models, regions, a cause or a workaround. We therefore cannot infer that every generation failed, that credits were lost, or that this was a recurrence of the earlier Seedance moderation event. Those are separate incident records.

Compliance API: cost-reporting backlog resolved on 12 October

Update, 13 October: OpenAI marked this event resolved at 19:02 Istanbul on 12 October. Original incident and recovery. The earlier monitoring notice had forecast catch-up through 11 October. This is a later recovery update, not a new outage. See the 13 October follow-up.

The original notices excluded ChatGPT usage and API Platform requests. Exact customer-impact boundaries remain unknown.

Earlier incidents: Claude reporting recovery added

Update, 14 October: Anthropic resolved this reporting incident on 13 October at 21:00 Istanbul. See the recovery roundup.

ElevenLabs’ Dubbing and SIP notices still show the recoveries covered in the 9 October incident edition; the SIP notice still promises a root-cause analysis. No fresh resolution of the earlier Suno, Runway credits, Moises lyrics or Kits download-minute customer reports was established. Isolated new complaints remain research leads rather than confirmed incidents.

Sources

• OpenAI: Android content, plugin administration, Compliance API Costs data.
• Runway: incident n70kzjr0x36k, public incident API.
• Anthropic: usage-data delay.
• ElevenLabs: Dubbing, SIP conversations.

Official notices establish acknowledged events; customer allegations require separate corroboration under our reporting method. Public-source coverage is selective. Google’s Gemini API status page and some review sources could not be verified in this scan; a coverage gap is not evidence of an outage or proof of normal service.

Update timeline

• 10 October 2026, 11:49 Istanbul / 08:49 UTC: Initial edition records four new official incident notices, including three provider-reported recoveries, plus the unchanged Anthropic reporting-delay follow-up. UTC timestamps are retained in the research ledger. No private account tests or paid generations were performed.

• 13 October 2026, 11:51 Istanbul / 08:51 UTC: Added the provider’s 12 October Compliance API recovery; original publication date and URL preserved.

• 14 October 2026, 13:26 Istanbul / 10:26 UTC: Added the later Anthropic reporting recovery; original publication date and URL preserved.