AI incidents: OpenAI Dots, ad-account and support recoveries — 23 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

OpenAI reports recovery from Dots message failures, new advertising-account administrator assignment failures and delayed support responses. These are three separate notices with different affected workflows. An additional community update reports success with an older business-icon upload problem; it does not establish a general recovery.

Research covers 20 October at 11:58 through 23 October at 11:57 Europe/Istanbul (UTC+3). Times below come from the original notices’ timestamped data, converted from UTC. Acknowledgement and resolution times measure the public incident timeline, not necessarily the full period of customer impact.

OpenAI: Dots message failures primarily affected Enterprise

OpenAI marked the Dots incident resolved at 23:55 Istanbul on 22 October. Users had been unable to send messages to Dots. The identification update says most impact was on Enterprise customers; it does not say that only Enterprise was affected. The listed component is ChatGPT Dots. Original Dots incident.

Investigation began at 21:51, followed by identification at 21:59. The approximately 2 hours 4 minutes to resolution is an acknowledgement-to-resolution interval. OpenAI gives no separate impact-start time, root cause, region breakdown or user workaround. This notice does not establish a failure of all ChatGPT messaging.

OpenAI Ads: new-account administrator assignment recovers overnight

New ad accounts were failing to assign their creator administrator access; OpenAI reports full recovery at 03:07 Istanbul on 23 October. The notice lists Ads Manager and Ads API. Original ad-account incident.

The first identified update was posted at 22:36 on 22 October. A further mitigation update followed at 00:30 on 23 October, then monitoring at 02:45. The full notice interval crosses midnight and lasts approximately 4 hours 31 minutes. It should not be read as a measured outage for every advertiser.

No root cause or account-specific repair instructions are supplied. The administrator-assignment failure is distinct from the earlier business-icon upload reports below; the status notice does not connect their causes.

OpenAI Support: response delays resolved after mitigation

OpenAI says its support-system disruption recovered at 08:58 Istanbul on 23 October. The first notice, at 06:41, warned of longer response times. At 06:57, OpenAI provided support@openai.com as a temporary contact route. A fix was being implemented by 07:43 and entered monitoring at 08:47. Original support incident.

The approximately 2 hours 17 minutes between investigation and resolution measures the notice interval. No product components are marked on the incident page, and no technical cause is disclosed. Recovery of the support system does not prove that every outstanding ticket has received a response.

Earlier ChatGPT reports: a limited Ads recovery and continuing scheduled-write failures

One community participant reports that the business-icon problem is fixed for their URL-based image upload. The post was published at 04:02 Istanbul on 23 October. It is a firsthand recovery claim, not an official announcement or confirmation that every upload method and account works. Recovery account.

We have added that dated qualification to the 19 October incident report, which covered the earlier staff acknowledgement and independent onboarding complaints. Its original publication date and historical cutoff remain intact.

Scheduled connected-app writes still have no confirmed general fix. A 22 October account describes a daily briefing generating content successfully while Gmail draft/send actions were blocked; interactive Gmail sends reportedly worked on the same account. This continues the earlier issue rather than establishing a new outage. Latest detailed account, earlier evidence and counterexamples.

Follow-up checks and practical next steps

Google’s London incident feed now aligns its recorded end with the preliminary report’s 15:10 Istanbul on 20 October main recovery boundary. The report remains preliminary; this is a record clarification, not another disruption or a new final root-cause report. Our 22 October coverage retains the distinction between main recovery and the later closure notice. Google record, timestamped feed.

Anthropic’s connector and Fable 5 events remain resolved in its incident history. ElevenLabs’ history provides no newly verified substantive incident for this window; promised analyses for earlier events remain unverified. These observations are limited to the sources we could inspect.

Affected users should match the named workflow to their failure, retain request IDs and timestamps, and check existing task or account state before repeating an operation. Provider recovery does not establish that a failed generation, message or account action completed automatically.

Sources

• OpenAI status: Dots, ad-account administrator assignment, support delays.
• OpenAI community: business-icon recovery account, scheduled Gmail-write account.
• Follow-ups: Google London record, Anthropic history, ElevenLabs history.

Public community posts are not necessarily verified customer reviews. We did not reproduce private-account failures, infer affected-user totals or treat inaccessible sources as evidence of good service. Isolated quality, credit and cancellation allegations remain unpublished leads. See the reporting method and separate product-news edition.

Update timeline

• 23 October 2026, 11:57 Istanbul / 08:57 UTC: Research cutoff. Added three official recoveries, a qualified Ads recovery account and follow-up checks. Added a dated note to the 19 October report.