OpenAI, the artificial intelligence research firm, restored full functionality to its suite of services late Monday following a significant global outage that disrupted access to ChatGPT, Codex, and its API platform. The incident, which occurred on April 20, 2026, left millions of users unable to access the company’s flagship AI tools during peak business hours in North America and Europe.
According to the official OpenAI status dashboard, the disruption began with localized issues affecting ChatGPT Business accounts overnight, specifically involving billing and the addition of new user seats. However, the situation escalated at approximately 10:05 a.m. ET (14:05 UTC), when the company acknowledged a broader partial outage. At the height of the disruption, Downdetector recorded over 8,700 reports of service failures in the United Kingdom and approximately 2,000 in the United States. Users in India and other regions also reported widespread Internal Server Error messages and total loss of access to both web and mobile interfaces.
The company confirmed that 12 separate components of the ChatGPT ecosystem were affected, along with the Codex application used by developers for automated programming tasks. While the Codex command-line interface (CLI) and VS Code extensions remained operational for some, the primary Codex App and the broader API platform—which powers thousands of third-party enterprise applications—experienced total downtime. OpenAI engineers applied a technical mitigation at 12:48 p.m. ET, after which services began to stabilize. By 1:45 p.m. ET, the company stated it was monitoring the recovery as traffic returned to baseline levels.
This infrastructure failure follows a period of unprecedented financial activity for the San Francisco-based company. Earlier in April 2026, OpenAI closed a landmark $122 billion funding round, which valued the firm at $852 billion—the highest valuation ever recorded for a private technology company. The company currently projects 2026 calendar-year revenue to exceed $30 billion, supported by a user base that has grown to approximately 910 million weekly active users.
The outage also coincided with a period of leadership transition within the company. Chief Operating Officer Brad Lightcap recently transitioned into a new role focused on special projects and private equity partnerships, with Denise Dresser assuming responsibilities as Chief Revenue Officer. Despite the disruption, OpenAI maintained its commitment to its enterprise Service Level Agreements (SLAs). The company has not yet released a formal post-mortem detailing the root cause of the database or server failures that led to the outage, though it noted that engineers are continuing to investigate the degraded performance to prevent future recurrences.