A global Salesforce outage disrupted customer access on Sept. 16, the second day of Dreamforce 2026, affecting logins, APIs, workflows, and partner-supported systems across multiple regions. According to Channel Insider reporting, customers experienced severe delays, intermittent errors, and service outages as engineers investigated issues with the company's login infrastructure and a core system component. The incident struck during Salesforce's flagship annual conference, one of its most important commercial events, and gave the disruption extra visibility as the company promoted its AI strategy.
Salesforce first acknowledged the problem at approximately 12:50 a.m. PDT, with multiple instances affected across all regions and hundreds of instances appearing on its Trust site. Requests stalled while waiting for a response from an internal login service, which consumed available server resources and prevented customers from creating new support cases during the incident. The company initially tested rolling restarts before reporting increased load on a core system component. By 3:56 a.m. PDT, Salesforce had validated a fix on a test instance, and a fleetwide rollout was underway by 4:19 a.m. PDT, with work continuing on a permanent code-level fix. The outage was separate from an interface-freezing problem affecting some Chrome and Edge 153 users, which stemmed from an external Chromium update rather than Salesforce infrastructure.
The disruption extended beyond employee login problems, with community reports of API integrations repeatedly triggering alarms. One user said their organization had to inform customers its system was unavailable and manage a growing backlog of callbacks. For Salesforce partners—consultants, integrators, and managed service providers—the incident created a familiar challenge: when a critical vendor goes down, customers often contact the partners managing their environments first, even when the underlying failure originates within Salesforce itself.
The report notes that the timing gave the incident extra visibility because it struck during Dreamforce, which this year is heavily focused on Salesforce's push into AI, including AIforce and an expanded Google Cloud partnership connecting Salesforce services to Google's AI infrastructure. Partners supporting business-critical Salesforce deployments may want to review three areas after an incident like this: integration monitoring to distinguish Salesforce service failures from issues in customer-built integrations, customer communication protocols to establish who updates customers when a vendor dependency interrupts a workflow or application, and fallback planning to identify processes that need a temporary alternative when Salesforce authentication or APIs are unavailable. The Sept. 16 outage lasted only a few hours, but it showed how a problem with login infrastructure can ripple through integrations and customer operations, serving as a reminder to build monitoring, fallback plans, and customer communication into deployments before the next service interruption puts those plans to the test. For partners connecting customer workflows, applications, and AI agents to the platform, reliability planning becomes even more critical as dependencies deepen and the surface area for cascading failures expands.