AI News Feed
Market watch
Products & Applications

Dreamforce 2026: Salesforce AI agents move from demos to measurable outcomes

At Dreamforce 2026, Salesforce and its ecosystem pushed AI agents beyond proof-of-concept demos, tying them to enterprise data, workflows, outcome-based pricing and task-level governance, SiliconANGLE reported.

The next phase of agent development is being defined by results rather than demonstrations, according to the report. Customers are seeking countable returns on investment, pricing tied to resolved issues, governance for each task and digital workers managed like employees. Those demands shaped Salesforce's latest agent announcements and its new AIforce layer.

George Gilbert, principal analyst for data platforms, intelligent apps and agentic frameworks at theCUBE Research, said Salesforce had assembled more of the pieces than other vendors. “What Salesforce has done probably more than any other vendor was put all the pieces together so that the customers don’t have to,” Gilbert said, according to SiliconANGLE. He added that this allows the industry to move beyond request and response to outcomes. Gilbert and co-host Alison Kosik interviewed industry experts during the San Francisco event, covering outcome-based pricing, digital labor, Slack as a platform for agents and the governance layer needed to scale agents safely.

Salesforce now charges for Agentforce service agents only when a customer’s issue is resolved, not when an agent hands the issue off to a human. Kishan Chetan, executive vice president and general manager of Agentforce and Service Cloud at Salesforce, said built-in analytics let companies keep tuning agents as their business processes change, according to SiliconANGLE.

At Asymbl Inc., digital workers make up half of the workforce across 13 functions. The company uses a digital recruiter that helped hire about 100 people in 100 days, and assigns each digital worker to a human manager with performance metrics. Brandon Metcalf, chief executive officer of Asymbl, said the productivity impact of those workers grew from about $5 million last year to about $13 million this year, SiliconANGLE reported.

Slackbot has become the front door for enterprise agents, according to the report. It now acts as a Model Context Protocol client that reaches every agent a user is permitted to use, pulls in trusted enterprise data and builds shareable interfaces on demand. Ryan Gavin, executive vice president and chief marketing officer of Slack at Salesforce, said Slackbot has become the fastest-adopted product in Salesforce history, with 1.5 million monthly active users since January.

Salesforce's Enterprise AI Harness combines Data 360, MuleSoft and Agentforce into an open, composable platform where agents inherit only the task-level policies their owners define. Savinay Berry, executive vice president and general manager of data management at Salesforce, said that granular control matters as chief information officers put governance ahead of demos and agents head toward creating 150 to 200 zettabytes of data, according to SiliconANGLE.