Dreamforce 2026: Salesforce bills Agentforce per resolved issue

Dreamforce 2026 marked a shift in how Salesforce's ecosystem talks about AI agents. The standalone toolkits that let technology teams build quick proofs of concept are giving way to integrated platforms that connect agents to enterprise data, to workflows and to the people who manage them. George Gilbert of theCUBE Research framed this as putting all the pieces together so customers don't have to, describing it as a harness that moves past request and response to outcomes.

The concrete examples at the event point in the same direction. Salesforce now bills Agentforce service agents on resolved issues, with built-in analytics for tuning as processes change. Asymbl runs digital workers across 13 functions, assigns each one a human manager and performance metrics, and says productivity from those workers grew from about $5 million last year to about $13 million this year, including a recruiter that helped hire about 100 people in 100 days. Slackbot has become a Model Context Protocol client inside Slack, reaching every agent a user is permitted to use.

Governance is the thread tying these together. Salesforce's Enterprise AI Harness gives agents only the task-level policies their owners define, which Savinay Berry tied to CIOs putting governance ahead of demos. The stakes may hinge on whether outcome-based pricing and task-level controls hold up as agents move further into daily work, and on whether buyers find the counting as credible as the demos were.