Salesforce just dropped a new version of Slackbot that actually does something useful. Not the old one that nagged you to add people to docs or archive channels. This one is a proper AI agent — it searches your company’s data, drafts docs, and can take actions on your behalf.
It’s now generally available for Business+ and Enterprise+ customers. And honestly, it’s about time Slack leaned into this properly.
Parker Harris, Salesforce co-founder and Slack’s CTO, didn’t mince words: “The old Slackbot was a little tricycle, and the new Slackbot is like a Porsche.” That’s marketing speak, but the technical changes are real. The old one was algorithmic and simple. The new one runs on an LLM with a robust search engine that hooks into Salesforce records, Google Drive, calendar data, and years of Slack conversations.
They kept the Slackbot name because people already know it. Fair call. But make no mistake — this is a complete rebuild.
What powers it? Anthropic’s Claude. Not because Salesforce loves Anthropic more than others — but because Slack’s commercial service has FedRAMP Moderate certification for U.S. government customers, and at the time, Anthropic was the only provider that could give them a compliant LLM. That exclusivity won’t last. Harris said they’ll support more providers this year, likely Google Gemini (he called it “incredible” with great performance and cost) and possibly OpenAI.
On the training data question — which always comes up — Harris was clear: Salesforce does not train any 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. Security and access control are real issues with LLMs, and Salesforce is smart to avoid that trap.
Salesforce tested this internally with all 80,000 employees. According to Slack’s CMO Ryan Gavin, it’s “the fastest adopted product in Salesforce history.” Two-thirds of employees tried it, 80% of those kept using it regularly, and internal 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 at the low end, it’s meaningful.
The adoption was mostly organic. Within five days, employees created a 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. People actually wanted to use this thing.
During a demo, Slack’s product experience designer Amy Bauer showed how Slackbot can synthesize info from multiple sources — customer feedback from a pilot program, a usage dashboard image, and Slack conversations — to produce executive-ready insights. That’s the kind of cross-source reasoning that old bots couldn’t touch.
Harris also echoed Marc Benioff’s view that LLMs are becoming commodities. “I call them CPUs.” That’s a good analogy — they’re the underlying compute, not the differentiator. The real value is in how you layer on enterprise data, security, and workflow integration.
One thing that bugs me: Salesforce is positioning this as the “front door to the agentic enterprise.” That’s a lot of buzzwords. But the product itself seems solid. The question is whether it can actually compete with Microsoft’s Copilot and Google’s Duet AI in the workplace. Those ecosystems are deeply entrenched. Slack’s advantage is its existing user base and the fact that it’s already a hub for internal communication. If Salesforce can make Slackbot truly indispensable — not just a nice-to-have — they might have a shot.
I’m keeping an eye on how third-party integrations evolve. If Slackbot can pull data from more sources without breaking security, it could become the central AI assistant for a lot of companies. For now, it’s a solid step forward — and a rare example of a major product overhaul that actually delivers on the promise.
Comments (0)
Login Log in to comment.
Be the first to comment!