Salesforce Suffers Global Outage During Dreamforce as AI Agent Pricing Shifts
Article excerpt
Highlighted: the sentence this signal was extracted from
A global software outage disrupted Salesforce customers worldwide on the second day of its flagship Dreamforce conference in San Francisco, compounding a week in which the company introduced a major shift in how it charges for artificial intelligence. As Salesforce rolled out new consumption-based and pay-per-resolution pricing models for its autonomous AI platform, Agentforce, the September 16, 2026, outage blocked logins and application programming interface (API) access for hundreds of instances globally. The simultaneous events highlight a growing tension for enterprise customers: businesses are becoming increasingly dependent on robust cloud infrastructure just as software vendors push them toward complex, value-based AI billing structures. The service interruption began at approximately 12:50 a.m. Pacific Daylight Time (PDT) and affected operating regions across the United States, Japan, India, France, and Germany. According to official incident reports, requests stalled while waiting for a response from an internal login service, which subsequently consumed available server resources and prevented users from accessing the platform. Initial attempts to resolve the issue through rolling restarts were unsuccessful, forcing engineers to develop a code-level fix. While some regions saw service restored by mid-morning, the outage persisted for up to nine hours in certain...
Keep reading with a free account
The rest of this article, and every signal for Salesforce, is in your free account.
