Salesforce (CRM) and Anthropic insist Claudeforce is about making workers better at their jobs, and they may be right.

That does not mean companies will need the same number of workers, though.

Claudeforce puts Salesforce data directly inside Claude, with a new Salesforce in Claude plugin shipping with 37 pre-built sales skills. Salesforce says those skills can analyze pipeline, flag deals at risk, build account plans, uncover renewals and expansion opportunities, prepare sellers for meetings, draft outreach, handle follow-ups and take actions inside the CRM.

Salesforce calls it “an AI CRO for every seller.”

Great for the seller. Potentially less great for everyone currently paid to do pieces of those 37 things.

That is where this stops looking like a productivity feature and starts looking like a headcount conversation.

A lot of corporate departments are ultimately collections of tasks. Sales operations analyzes pipeline, enablement prepares reps, analysts assemble reports, SDR teams research prospects and draft outreach, managers decide where attention should go, and CRM administrators help keep the underlying system updated.

Claudeforce does not eliminate those jobs with the flip of a switch. But start moving enough of those tasks into one agent and the headcount math gets pretty difficult to ignore.

Salesforce itself says a seller can ask Claude what needs attention that morning and receive an action plan based on live pipeline data. It can generate plans for individual deals, surface growth opportunities, prepare someone for a meeting and handle the follow-up.

And it can act, not just recommend.

That means disruption.

A five-person team becomes four. Someone leaves and isn’t replaced. One analyst supports twice as many sellers. A manager covers a larger organization. Eventually, a company realizes it is producing the same amount of work with fewer people.

The department still exists. There are just fewer chairs in it.

Anthropic CEO Dario Amodei addressed the broader fear in an interview with CNBC’s Jim Cramer.

“We’re not interested in destroying anyone,” Amodei said, describing AI as a “positive sum” technology.

Fair enough.

But his next point may be the more important one.

“It’s about how much of these enormous gains go to various people and various companies,” he said, adding that Anthropic wants its customers to “share these gains with us.”

There it is.

If AI creates enormous productivity gains, somebody gets them. Employees could get higher wages or shorter workweeks. Customers could get lower prices. Companies could produce more.

Or shareholders could get higher revenue per employee and wider margins while the org chart gets a little shorter.

And that is what makes Claudeforce more interesting than another AI product launch. Salesforce and Anthropic have packaged 37 pieces of knowledge work into software that can reason across a company’s data, decide what needs to happen and then execute against it.

Nobody has to set out to destroy a department.

You just have to make enough of the work inside it unnecessary.

I guess we’ll see.