Salesforce's Dreamforce 2026 strategy fundamentally reframes CRM as infrastructure rather than application. The company is embedding customer data, business rules and approved actions into the tools where work actually happens—Slack, Claude, Microsoft Teams—rather than requiring employees to navigate back to the CRM interface itself. This shift manifests through three interconnected components: ClaudeForce integrates Claude Enterprise with Salesforce data and security controls; Slackforce Surfaces creates interactive, live-updating dashboards within Slack channels; and a portfolio of prebuilt agents (Casey for service, Fin for complex cross-channel workflows) handles customer interactions autonomously across voice, SMS, WhatsApp and web. The architectural ambition is substantial—Salesforce is positioning CRM as the governed layer beneath customer operations, whether work occurs in a service application, a Slack channel, an AI assistant or a connected third-party system.
For CX teams, the appeal is immediate but the operational complexity is real. Reducing context-switching by surfacing customer information, SLA breaches and case severity directly in Slack could accelerate response times and improve queue visibility. Yet this distributed model introduces governance questions that existing CRM workflows largely avoided. When an employee or AI agent acts from Slack or Claude rather than within Salesforce's native interface, which customer data informed the decision? Which policy applied? Can the organization audit and correct the action if something goes wrong? Salesforce's answer—the Trusted Enterprise AI Harness and a new AI Control Plane—promises to manage agents across Salesforce and third-party environments, but the company has not yet demonstrated these elements working together in real customer journeys. The underlying technologies exist; unified experience and additional capabilities are expected in early fiscal 2028, with pricing unannounced.
The strategic implication cuts deeper than interface design. For teams already running Agentforce or considering deployment, this shift means governance and auditability become as critical as automation rates. Salesforce's claims that Casey and Fin resolve high percentages of interactions autonomously are useful deployment benchmarks, but they obscure the harder metrics: repeat contact, abandonment, complaint rates, customer effort and escalation quality. CX leaders must establish what governance actually looks like when customer interactions happen outside the CRM—how permissions apply when data from multiple systems flows into Slack, whether live operational views are current and complete, and whether the organization can trace decisions made by distributed agents. The architectural shift also raises questions about vendor lock-in and integration depth. Microsoft Teams is named as part of the strategy, but Salesforce has not clarified whether that integration will match Slack's depth or remain a lighter-touch connection. For teams evaluating CRM strategy, the question is not whether to adopt this model, but whether your organization's governance maturity and technical infrastructure can support it.
SAN FRANCISCO — Salesforce has arrived at Dreamforce 2026 with an argument that would once have sounded unusual for a CRM company: the future of CRM may involve spending less time inside the CRM itself. The company wants customer data, business rules and approved actions to be available in the tools