Skip to content
Back to OpenAIIncident Postmortem
OpenAI logo

OpenAI

Tracked by PulseWatch

Auto-drafted postmortem

OpenAI outage on May 4 — 33-minute disruption

May 4, 10:15 AM → May 4, 10:49 AM · 33 minutes · high

Users experienced a service disruption with OpenAI, which was detected through user reports. The incident, characterized by slow or unresponsive service, began at 10:15 UTC and was resolved at 10:49 UTC on May 4, 2026.

Peak reports
10
in 15 min
Total reports
12
during the incident
Duration
33 minutes
tracked by PulseWatch

Impact

The primary impact observed by users was slow or unresponsive service across the OpenAI platform. This was reflected in 12 user reports received during the incident, with a peak of 10 reports within a 15-minute window. No specific geographic regions were identified as disproportionately affected, indicating a general service degradation. The disruption lasted for approximately 33 minutes.

Top reported issues

  • Slow / unresponsive12

Incident timeline

  1. May 4, 10:15 AM

    PulseWatch detected elevated report volume (high severity)

  2. May 4, 10:30 AM

    Peak of 4 reports in 15 minutes (baseline 1.3)

  3. May 4, 10:43 AM

    Status updated: possible issue

  4. May 4, 10:49 AM

    Report volume returned to baseline — incident resolved

Official updates

  1. May 4, 10:15 AM

    OpenAI status page acknowledged the incident.

    View official source
Draft auto-generated by PulseWatch when the incident resolved. Based on aggregated user-report signals and the sources listed above.
This postmortem was auto-generated from PulseWatch's record of the incident. · Model: gemini-2.5-flash