Claudeforce: The #1 AI Meets the #1 CRM

For decades, software gave us fixed interfaces: screens, menus, workflows designed in advance. The information was in there, but assembling it was your job — open the list, click into the account, read back through the history, hold it all in your head, and only then decide what to do. A seller could lose a morning before making a single call.

Claude changes that. It reasons across your systems, reaching into your data, bringing the right answer and the right action instead of making you navigate to find it. This is a shift from software as the interface to software powering every interface.

A model can reason brilliantly and still be a genius who's never seen your deals. Point it at your company and it can't tell a healthy deal from one about to slip, because it doesn’t know your pipeline, your accounts, or your history. That's what Salesforce in Claude is all about. Turn the connection on once, and Claude can reach across Salesforce, Slack, and your email to reason over the real state of your business and act on it.

This past year, Salesforce and Anthropic have built together again and again, from Claude Tag in Slack to Claude in Agentforce, and most recently Claude as a launch partner for Slack Code, the space where engineering teams and agents build in the open. Each new capability had the same goal: put AI in the tools people already use, for engineers and knowledge workers alike. Claudeforce is bigger than any one of them. It's the moment the #1 AI and the #1 CRM start working as one, giving every seller an AI CRO.

Viewing your pipeline is one thing, and plenty of tools can do it. But acting on it is harder, because it means updating records, running multi-step workflows, and moving a deal forward without stepping outside the rules your company enforces. No two companies run sales exactly alike, which is why Claude routes every action through Salesforce, so your rules are enforced.

With Claudeforce, we are building on Headless 360, a collection of Salesforce capabilities (data, apps, workflows, agents, and the governance around them) that AI can call directly over MCP, no user interface required and no custom integration work. For years that value was locked behind the interface: it lived in your data, your relationships, and your rules, but the only way to reach it was to log in and click through screens. MCP is what frees it, so an AI agent can work with it wherever you are. The same capabilities are open to any developer who wants to build their own, but through Claudeforce, you get a ready-made experience, built for sellers, that works out of the box.

Every answer and every action runs through your existing Salesforce permissions and business rules. Claude sees what the user is authorized to see, and can do what that user is authorized to do, no more. There's nothing new to stand up, nothing to re-audit, and nothing to configure account by account. An admin connects Salesforce in Claude once, and it works for the whole team from day one.

On the write side, the controls are just as concrete, and you decide how they work. Claude can check with you before it emails anyone outside your company, or just send it once you're comfortable letting it. When it updates a record, it changes only the field it said it would. Humans direct. Agents execute. Salesforce governs.

Salesforce worked with Anthropic on Enterprise Frontier Safeguards, which combine the privacy of zero data retention with state-of-the-art safeguards for detecting serious misuse. Eligible customers will be able to keep activity data in cloud infrastructure they control, under their own encryption keys, access policies, and audit logging. When monitoring detects a pattern that needs attention, the alert goes directly to the customer for review, with no Anthropic human review required.

Enterprise Frontier Safeguards will roll out in phases beginning later this fall 2026. Read Anthropic’s announcement and request access .

Join us on September 15–17, 2026 in San Francisco.