Slackbot's 30 AI Features: The Most Ambitious Enterprise Agent Push Since Salesforce

The Big Picture

Slack today announced more than 30 new capabilities for Slackbot at a keynote headlined by Salesforce CEO Marc Benioff. Less than three months after Slackbot became generally available in January, Slack is calling it the fastest-adopted product in Salesforce's 27-year history — with some enterprise users reporting up to 90 minutes of daily time savings. Inside Salesforce, teams claim up to 20 hours per week.

Rob Seaman, Slack's interim CEO and former chief product officer, put it plainly: "The upper bound of use cases is effectively limitless." That's a significant claim from someone running the show at one of the world's largest enterprise software companies.

What Actually Shipped

The update organizes around several major capability areas:

• AI-Skills — Reusable instruction sets that define inputs, steps, and output formats for repeatable tasks. Teams can build a skill once and deploy it on demand. Slackbot recognizes when a prompt matches an existing skill and applies it automatically.
• Deep Research Mode — Gives Slackbot the ability to run extended, multi-step investigations that take approximately four minutes to complete. Slack deliberately didn't demo this on stage — the value is in depth, not speed.
• MCP Client Integration — Slackbot can now call tools in external systems through the Model Context Protocol. That means it can create Google Slides, draft Google Docs, and interact with the 2,600+ apps in the Slack Marketplace and 6,000+ apps on Salesforce AppExchange. "We're going all in on MCP for Slackbot," Seaman said.
• Meeting Intelligence — Listens to any meeting — not just Slack Huddles, but Zoom, Google Meet, or any provider — by tapping into local audio through the desktop app. Captures discussions, summarizes decisions, surfaces action items, and logs directly to Salesforce CRM.
• Slackbot on Desktop — Extends the agent outside the Slack application window entirely.
• Voice Mode — Text-to-speech and speech-to-text available now; full speech-to-speech under active development.

Claude Under the Hood — and the Cost Problem

Slackbot is built on Anthropic's Claude model. The partnership was confirmed ahead of the keynote, where Anthropic leadership also appeared on stage alongside Slack executives.

Managing the cost of Claude's reasoning at enterprise scale is one of the most significant technical challenges the team faces. Slackbot is included in Business+ and Enterprise+ plans at no additional charge — meaning Slack carries the cost optimization burden, not customers.

"A lot of what we've done is in the context engineering phase, working really closely with Anthropic to make sure that we're optimizing the RAG phase, optimizing our system prompts and everything, to make sure we're getting the right amount of context into the context window and not obviously making fiscally irresponsible decisions for ourselves," Seaman told VentureBeat.

Starting in April, Slackbot will also roll out in limited capacity to free and Pro users — a deliberate funnel move to drive upgrades.

The Surveillance Question

The extension of Slackbot beyond the Slack application — particularly its ability to listen to meetings and view screen content — raises immediate questions about workplace surveillance. In large enterprises where tens of thousands of workers operate under company-wide IT policies, these capabilities could be turned into broad monitoring tools.

Seaman said every capability is opt-in at the individual level and that enterprise admins have separate controls. Whether that satisfies corporate legal teams and labor regulators remains to be seen — this is uncharted territory for workplace privacy law.