OpenAI's ChatGPT conversation-error incident closes after about 42 hours with no root cause published

Sources: OpenAI status, HN 49046049
Summary: The incident opened 2026-07-25 22:09 UTC covering intermittent errors loading or continuing ChatGPT conversations, with dated impact from about 13:00 PT. It was identified at 23:16 UTC and moved to monitoring at 23:57 UTC with a note that mitigation had been implemented. It then sat in monitoring for over a day and resolved 2026-07-27 16:32 UTC with the update that all impacted services had fully recovered, about 42 hours after it opened. No root cause is published. Earlier runs of this digest recorded the incident at 15 hours and at 29 hours open while it was still in monitoring. Two other 2026-07-25 incidents on the same page, both titled Elevated error rates, were resolved earlier that day, at 11:08 UTC and at 11:57 UTC.
Why it matters: A mitigation that was neither confirmed nor withdrawn for a day, then closed with no explanation, gives teams on the ChatGPT surface no basis to size retry and fallback behaviour against a recurrence.
Follow-up: Watch for a root-cause note or incident summary covering the 2026-07-25 to 2026-07-27 window.