Salesforce just dropped a massive AI overhaul for Slack, packing 30 new features into the enterprise messaging platform. CEO Marc Benioff and his team unveiled the updates at a small gathering in San Francisco on Tuesday, as TechCrunch AI reports.
The headline feature? A seriously upgraded Slackbot that’s gone from a basic assistant to a full-blown AI agent with what Salesforce calls “reusable AI-skills.” These let users define specific tasks that, once created, work across different scenarios and contexts. Think of them as programmable shortcuts for your AI assistant.
What Slackbot Can Actually Do Now
- Reusable AI-skills: Define a task once, use it everywhere. Slackbot ships with a built-in library, but you can also create custom skills. Trigger them with simple commands like “create a budget” and the bot pulls data from Slack channels, connected apps, and other sources to build an actionable plan.
- MCP client support: Slackbot now speaks Model Context Protocol, meaning it can connect to and coordinate with outside services and tools, including Salesforce’s own Agentforce platform.
- Meeting transcription and summaries: Zoned out during a call? Ask Slackbot for a recap, complete with action items assigned to you.
- Desktop monitoring: The agent can now operate outside Slack itself, watching your deals, conversations, calendar, and habits to make proactive suggestions and draft follow-ups.
- Agentforce integration: Through MCP, Slackbot routes work and questions to Agentforce or any enterprise agent, finding the most efficient path without human intervention.
The Bigger Picture
This update follows a January release that first gave Slackbot agentic capabilities like drafting emails, scheduling meetings, and inbox search. The 30 new features will roll out “in the coming months,” according to TechCrunch AI.
What stands out here is how aggressively Salesforce is pushing Slack beyond messaging. The desktop monitoring feature is particularly ambitious. Salesforce says privacy protections are built in and users can adjust permissions, but expect this to raise eyebrows in enterprise security circles. Rob Seaman, Slack’s interim CEO, emphasized the privacy-first design.
The strategic play is clear: Salesforce wants Slack to become the central operating layer for enterprise work, not just the place where you send messages and share memes. By embedding AI deeply into workflows, pulling data from across an organization, and connecting to external tools via MCP, they’re positioning Slack as an indispensable business platform.
Salesforce’s Slack Bet by the Numbers
Benioff shared some context during his keynote: five years after acquiring Slack, the platform has delivered “two and a half times revenue growth” with about a million businesses running on it. That’s a strong foundation, but the AI features are clearly designed to justify Slack’s place in an increasingly competitive enterprise AI landscape.
The timing matters too. With Microsoft pushing Copilot deep into Teams and Google embedding Gemini across Workspace, Salesforce needs Slack to be more than a chat app. The MCP integration is a smart move. It positions Slackbot as a universal connector rather than a walled garden, which could be a real differentiator.
All 30 features are expected to become available in the coming months. For the full details, check out the original reporting from TechCrunch AI.