Salesforce’s New Slackbot Is a Porsche, Not a Tricycle: Inside the AI Rebuild

Salesforce just dropped a fully rebuilt Slackbot, and it’s not your grandpa’s notification bot anymore. The old version did simple stuff—remind you to tag someone, suggest archiving a channel, maybe deliver a “you got a message” alert. Boring, but functional.

The new Slackbot is an AI agent. It can search your Salesforce records, dig through Google Drive, pull calendar data, and sift through years of Slack conversations. It writes drafts, summarizes customer feedback, and takes actions on your behalf. Parker Harris, Salesforce co-founder and Slack’s CTO, put it bluntly: “The old Slackbot was, you know, a little tricycle, and the new Slackbot is like, you know, a Porsche.”

I’ve seen this kind of language before—execs love car metaphors—but the underlying shift is real. The architecture is entirely new: built around an LLM and a robust search engine, with connections to third-party enterprise data sources. They kept the Slackbot name because, as Harris said, “People know what Slackbot is.” Fair enough. Brand recognition matters.

Why Anthropic’s Claude and What About Other Models?

The new Slackbot runs on Claude from Anthropic. That choice wasn’t purely technical—it was driven by compliance. Slack’s commercial service has FedRAMP Moderate certification for U.S. federal customers, and Harris said Anthropic was “the only provider that could give us a compliant LLM” when they started building.

But exclusivity won’t last. Harris told me they plan to support additional providers this year. “We have a great relationship with Google. Gemini is incredible—performance is great, cost is great. So we’re going to use Gemini for some things.” OpenAI is also on the table. He echoed Marc Benioff’s line that LLMs are becoming commodities: “I call them CPUs.”

I’m not fully sold on the commodity argument—model quality still varies a lot—but the multi-model approach makes sense for enterprise customers who want flexibility or have existing relationships.

The Data Question: No Training on Customer Conversations

One concern with any AI agent is data privacy. Harris was clear: Salesforce doesn’t train models on customer data. “Models don’t have any sort of security,” he explained. “If we trained it on some confidential conversation that you and I have, I don’t want Carolyn to know—if I train it into the LLM, there is no way for me to say you get to see the answer, but Carolyn doesn’t.”

That’s a solid stance. It avoids the kind of data leakage nightmares we’ve seen with other AI products. But it also means the bot can’t learn from your specific conversations to get smarter over time. Trade-offs.

Internal Results: 80,000 Employees, 96% Satisfaction, 20 Hours Saved

Salesforce tested this thing internally for months with all 80,000 employees. According to Ryan Gavin, Slack’s CMO, it’s “the fastest adopted product in Salesforce history.” Two-thirds of employees tried it, and 80% of those kept using it regularly. Satisfaction hit 96%—the highest for any AI feature Slack has shipped.

Employees reported saving between two and 20 hours per week. That’s a huge range, but even on the low end, it’s meaningful. Adoption was largely organic. Within five days, employees created a shared Canvas called “The Most Stealable Slackbot Prompts” that now has over 250 prompts. Kate Crotty, a principal UX researcher, found that 73% of internal adoption came from social sharing, not top-down mandates.

That kind of organic virality is rare in enterprise software. It suggests the bot actually solves real problems.

What It Can Actually Do

During a demo, Amy Bauer, Slack’s product experience designer, showed how Slackbot can synthesize information from multiple sources. She asked it to analyze customer feedback from a pilot program, upload an image of a usage dashboard, and correlate the data. It handled the whole thing in natural language.

The bot can also take actions: update Salesforce records, create tasks, schedule meetings. It’s not just a query tool—it’s an agent that does things. That’s the “agentic AI” pitch Salesforce has been pushing, and it seems to hold up in practice.

The Competitive Landscape

Salesforce is fighting Microsoft and Google for workplace AI dominance. Microsoft has Copilot across its stack. Google has Gemini in Workspace. Slackbot is Salesforce’s counterpunch, and it’s a strong one—especially for companies already deep in the Salesforce ecosystem.

But the real test is whether customers will pay for it. The new Slackbot is generally available to Business+ and Enterprise+ customers. That’s a premium tier, and adoption will depend on whether the ROI is visible. Early internal numbers are promising, but enterprise sales cycles are long.

I’m cautiously optimistic. The rebuild is significant, the privacy stance is responsible, and the organic adoption suggests genuine value. But I’ve seen plenty of enterprise AI products that looked good in demos and flopped in the real world. We’ll know more in six months.