ChatGPT is having a partial outage

OpenAI experienced a prolonged partial outage on Tuesday, impacting access to ChatGPT, Sora, and its API, with users experiencing elevated errors and latency; the outage, which began late Monday, persisted into Tuesday morning and may take several more hours to fully resolve. This disruption coincides with a flurry of recent announcements, including deepened integrations with Apple and confirmation of $10 billion in annualized recurring revenue, as well as CEO Sam Altman's announcement of an 80% price cut for developers accessing its o3 AI reasoning models. The outage highlights the challenges OpenAI faces in scaling its infrastructure to meet rapidly increasing demand, as Altman previously noted the company's GPUs are struggling to keep pace.

Analysis

OpenAI is experiencing a significant partial outage across ChatGPT, Sora, and its API, which began late Monday and persisted into Tuesday morning, with full service restoration expected to take several more hours. This disruption, marked by user reports of 'elevated errors and latency' and 'Too many concurrent requests' for GPT-4o, is notably prolonged. The outage coincides with a period of strong positive momentum for OpenAI, including the announcement of $10 billion in annualized recurring revenue, an 80% price cut for its o3 AI reasoning models via API for developers, and deepened integrations with Apple showcased at WWDC. These events underscore the immense demand for OpenAI's services and the concurrent operational strain, as CEO Sam Altman previously highlighted the company's 'GPUs are melting' in efforts to scale, indicating that computing resources remain a critical bottleneck.

Key Decisions for Investors

• Investors should note the operational risks inherent in rapidly scaling AI platforms, as evidenced by OpenAI's prolonged outage, which could affect user trust and the businesses reliant on its services, despite strong financial metrics like its $10 billion annualized recurring revenue.
• For companies like Apple (AAPL) pursuing deeper integrations with OpenAI, this incident highlights potential third-party operational risks that could impact the reliability of new product features and warrants careful due diligence on partner infrastructure stability.
• The persistent compute constraints and service instability across leading AI providers underscore the significant ongoing capital expenditure required in the AI sector, potentially benefiting GPU manufacturers and cloud infrastructure providers, while also creating openings for AI competitors demonstrating more robust scaling solutions.