AI incidents: OpenAI Codex, ElevenLabs Agents and Claude reporting recoveries — 14 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 summary and status at publication

Four official recovery records qualify today, including two distinct OpenAI incidents. The scan covers 11 October at 11:51 through 14 October at 13:26 Europe/Istanbul (UTC+3). The edition includes later developments in previously reported events; it does not imply every disruption began today.

These are announcement times, not measured impact boundaries. See the separate 14 October AI news edition for product releases and updates.

OpenAI: today’s Codex and Work errors have recovered

OpenAI marked its 14 October degradation resolved at 12:53 Istanbul. The affected components are ChatGPT Work and Codex in ChatGPT Desktop. It acknowledged elevated errors at 10:05, then said mitigation was still being implemented at 12:19. Official incident and updates.

The notice supplies no model, regional breakdown, root cause, precise impact start or customer workaround. It does not establish that all ChatGPT or OpenAI API products failed. The interval between acknowledgement and closure is therefore not presented as an outage duration.

Yesterday’s GPT-6.1 Sol event is a separate recovery

OpenAI closed the 13 October GPT-6.1 Sol incident at 19:36 Istanbul, after moving to monitoring at 12:30. That record lists ChatGPT Work, Codex Web, Codex API, CLI and the VS Code extension. Yesterday’s incident.

We have updated the 13 October edition in place. The two incident IDs do not establish a common cause or a continuous disruption. Exact customer-impact boundaries remain unspecified.

ElevenLabs: US Agents recovered after repeated database disruption

ElevenLabs’ 13 October notice records recurring latency, timeouts and failed calls, followed by recovery. Its affected-component list names ElevenAgents conversations, telephony, RAG, quality, other API endpoints, UI and integrations. Original incident.

Provider explanation and timeline limits

ElevenLabs attributes the US disruption to its MongoDB-backed Agents database following maintenance. Its incident write-up describes write stalls and repeated failovers, followed by stabilization and added capacity. It lists further configuration, monitoring and architecture work as follow-up plans, not completed guarantees.

The write-up’s detailed clock times have no stated timezone. We preserve the timezone-labelled public notices above and do not silently convert the write-up’s times or infer a single outage duration. Its publication time is also unstated; it was accessible during this scan, although a later retrieval failed. The evidence concerns the described US Agents incident, not a confirmed outage of all ElevenMusic, TTS or video generation.

Anthropic: the earlier usage-reporting delay is resolved

Anthropic marked the Console reporting issue resolved on 13 October at 21:00 Istanbul. Core API requests were excluded from the reported impact. Official recovery.

The 8 October report and later follow-ups now carry the recovery. No precise impact interval is established.

What affected users can check

Use the linked incident matching the service and date you used. None of these notices establishes automatic replay of failed work, refunds or recovery of every account. If a problem persists after provider resolution, retain the affected component, timestamp and error when contacting that provider’s support; keep credentials and private content out of public reports.

The earlier Firefly Video color reports and ChatGPT loading/capacity accounts have no newly established provider fix in the checked discussions. They remain separate customer reports, with no demonstrated connection to today’s events.

Sources

• OpenAI: 14 October desktop/Work recovery and 13 October GPT-6.1 Sol recovery.
• ElevenLabs: incident history and notices and provider write-up.
• Anthropic: Console reporting incident.

Music, video and remix/stem reviews were checked alongside major-provider histories. Isolated new billing and output-quality allegations did not meet the corroboration threshold. Gemini API history, some review pages and several incomplete status histories remain coverage gaps. Absence of a qualifying item proves neither normal service nor disruption. See our reporting method.

Update timeline

• 14 October 2026, 13:26 Istanbul / 10:26 UTC: Initial edition records four provider recoveries. The 8, 9, 10, 11 and 13 October incident articles receive dated recovery updates. No private account tests or paid generations were performed.