Salesforce’s Summer ‘26 release, available June 15, marks a turning point in how enterprise AI platforms are measured. The revenue numbers are no longer speculative — Agentforce annual recurring revenue reached $800 million, up 169% year over year, with combined AI revenue surpassing $2.9 billion. The company closed 29,000 Agentforce deals in the last year, up 50% quarter-on-quarter in Q4 alone, and logged 2.4 billion agentic work units across Agentforce and Slack.

These are not pilot numbers. They are production numbers.

What’s Actually Shipping on June 15

The headline feature is multi-agent orchestration. It works like this: one primary agent receives the user request, routes it to the best-fit specialist agent using Salesforce’s Atlas 3.0 reasoning engine, and returns a coherent, contextual answer without the user needing to switch interfaces, re-explain context, or manage which agent handles which task.

The underlying infrastructure for this is more interesting than the UX. Atlas 3.0 now supports the Agent-to-Agent (A2A) protocol — the open standard Google and others are pushing to enable agents from different vendors to hand off work reliably. Alongside that, Tableau is now accessible via the Model Context Protocol (MCP), which means agents can pull live analytics data into their responses without requiring a custom integration layer.

For businesses already running Salesforce, this matters practically: a service agent can now hand a complex request to a finance agent, which can pull a Tableau dashboard, summarise it, and return results inside the same conversation thread. That kind of cross-domain, multi-step workflow was impossible in a single-agent architecture.

Other notable features in the Summer ‘26 release:

IT Service Domain Pack — ships with 50 pre-built AI agents covering IT service management workflows, intended to reduce the setup time from months to days for standard enterprise IT processes.

Slack First Sales — brings Agentforce into Slack, with agents that prospect leads, update pipeline, and surface deal context where sales teams already spend their time. Salesforce is clearly trying to reclaim Slack as the centre of enterprise work rather than letting Microsoft Teams hold that position.

Process Compliance Navigator — monitors live workflows in real time and intercepts non-compliant actions before they complete. For finance, legal, and healthcare businesses under regulatory pressure, this is the feature that turns AI agents from a productivity tool into a compliance tool.

Why the Revenue Numbers Matter

The $800M ARR figure is worth pausing on. Salesforce launched Agentforce in early 2025. Reaching $800M ARR in roughly 18 months is one of the fastest enterprise software growth runs in recent history. For context, Salesforce took years to reach similar milestones with its core CRM product.

What it signals to the broader market: enterprises are not just evaluating AI agents. They are paying for them at scale. The 29,000 deal count means Agentforce has gone well beyond the Fortune 500 early adopter phase into mid-market territory. That is the same adoption pattern that defined SaaS in the early 2010s.

The 2.4 billion agentic work units logged is a different kind of number — it measures actual usage rather than sold seats. Each unit represents a task an agent completed autonomously. At that volume, Salesforce can start making statistically meaningful claims about what agents actually do, how often they succeed, and where they fail. No other enterprise AI platform has disclosed a comparable usage volume.

What This Means for Business

Single agents are no longer the benchmark. Businesses that deployed their first AI agent in 2025 thinking that was the destination are now behind. The question is no longer “do we have an agent?” It is “how do we orchestrate multiple agents across departments?”

MCP is becoming the connective tissue. Tableau joining the MCP ecosystem follows Workday, Microsoft, Google, and others. The practical takeaway: if you are building an AI agent strategy now, any tool that does not support MCP is either going to add that support quickly or be replaced by one that does. Building agent workflows around MCP-compatible tools reduces fragmentation as the ecosystem matures.

Compliance is becoming an agent feature, not a separate process. The Process Compliance Navigator is a signal that regulated industries — financial services, healthcare, legal — are now the primary growth market for enterprise AI agents. These sectors held back from early agent deployments because agents that autonomously take actions create audit problems. Agents that intercept non-compliant actions before they complete solve that problem natively.

Revenue growth at this rate forces a strategy question. If Agentforce can grow from near zero to $800M ARR in 18 months, businesses that wait another 18 months to start building their agent workflows are conceding two cycles of productivity gains to competitors who did not wait. The question is no longer whether agent deployment is worth the investment. The scale of adoption is answering that.

---

Enterprise DNA’s Omni Advisory service helps business leaders design multi-agent workflows that integrate with the platforms your teams already use — including Salesforce, Workday, and the broader agentic ecosystem.