Slack's AI agent just got a serious promotion
If you use Slack for work, your experience is about to change significantly. Salesforce CEO Marc Benioff unveiled 30 new AI features for Slack, making it the most ambitious update the platform has received since Salesforce acquired it in 2021 for $27.7 billion. The centerpiece is Slackbot — Slack's built-in AI assistant — which has been transformed from a basic bot into something much more capable.
Back in January, Slackbot got its first major AI upgrade: the ability to draft emails, schedule meetings, and search your inbox for specific information. That was the foundation. What got announced now builds on that foundation substantially — and some of it is genuinely impressive.
Reusable AI skills — the most interesting new feature
The feature that stands out most is what Salesforce calls "reusable AI skills." Here's the concept: instead of describing a complex task to an AI every single time you need it done, you define the task once — create a reusable skill — and then trigger it with a simple command whenever you need it.
Here's a concrete example. Say you regularly need to put together budget plans for events or projects. Instead of manually pulling information from multiple Slack channels, Google Drive, Salesforce CRM, and other tools every time, you define a skill called "create a budget." Now whenever you type that command in Slack, Slackbot automatically pulls together all relevant information from across your connected apps and channels, assembles an actionable budget plan, and — without you asking — sets up a meeting with the relevant team members (identified by their job titles) to review it.
That last part is key: the meeting gets scheduled automatically, with the right people invited, without you doing anything beyond triggering the skill. For teams that run the same types of projects repeatedly — which is most teams — this kind of automation removes genuinely tedious work.
Slackbot comes with a library of pre-built skills you can use immediately, but you can also create completely custom skills tuned to your specific workflows and company processes.
MCP integration — Slackbot connects to everything
Slackbot now functions as an MCP (Model Context Protocol) client. MCP is an open standard developed by Anthropic that lets AI tools connect to and coordinate with external services and data sources using a common language. Think of it like a USB standard for AI integrations — once a service supports MCP, any MCP-compatible AI tool can connect to it without custom integration work.
For Slack users, this means Slackbot can now communicate with and delegate tasks to other AI agents and enterprise tools outside of Slack. The most immediate integration is with Agentforce — Salesforce's AI agent platform launched in 2024. Through this connection, Slackbot can route work to Agentforce agents automatically, letting Agentforce handle specialised tasks (customer service workflows, sales data analysis) while Slackbot coordinates the overall process — without requiring human direction at each step.
Meeting transcription and real-time recaps
This one is practical and immediately useful. Slackbot can now transcribe meetings and generate summaries in real time. If you zone out during a call, lose track of a discussion, or join late, you can ask Slackbot for a catch-up recap — including any action items that were assigned to you specifically.
The meeting intelligence capability works across different video call providers, not just Slack's own Huddles. This is important because most enterprise teams use a mix of Zoom, Google Meet, Microsoft Teams, and Slack Huddles depending on the context.
Desktop monitoring — the privacy conversation nobody's having
The most eyebrow-raising feature is this: Slackbot can now monitor your desktop activity outside of Slack. Salesforce says it tracks "your deals, your conversations, your calendar, and your habits" to make proactive suggestions and draft follow-ups for critical tasks.
The company says privacy protections are built in and users can adjust what the bot monitors. But let's be honest — an AI assistant watching your desktop activity is a significant permission to grant any software. For employees using company-issued devices where IT has visibility anyway, this might be less concerning. For individuals using personal devices for work, it's worth reading the fine print on exactly what gets monitored and what data leaves your device.
Slack's interim CEO Rob Seaman emphasized that the goal is to remove cognitive overhead — instead of you having to remember to follow up on a deal or check a calendar conflict, the system notices and acts. Whether that feels helpful or intrusive will depend heavily on how it's implemented and what controls users actually have.
What this means for Indian IT professionals and enterprises
Slack is widely used across Indian IT services companies, startups, and enterprise teams. TCS, Infosys, Wipro, and hundreds of Indian product startups use Slack as their primary communication platform. These 30 new features will roll out over the coming months to paid Slack plans.
For Indian IT professionals working in offshore delivery teams, the meeting transcription and recap features are particularly valuable. Cross-timezone communication is a constant challenge — US clients finish meetings that Indian teams join late, or Indian teams have sync calls at awkward hours. Having AI-generated summaries and action item lists removes some of the timezone friction.
The reusable AI skills are most valuable for teams running repetitive project workflows — which is essentially every delivery team in an IT services company. If you can automate the "pull together status updates, draft a summary, send to the client" workflow with a single command, that's hours saved per week across a team.
Salesforce's Agentforce integration also matters for Indian companies using Salesforce CRM — which includes a significant portion of enterprise India. The ability to trigger CRM workflows from a simple Slack command, without switching applications or asking an admin, reduces one of the most common friction points in enterprise software use.
Microsoft Teams is watching nervously
Slack's direct competitor in enterprise messaging is Microsoft Teams. Teams is deeply embedded in Indian enterprise through Microsoft 365 bundles. But Microsoft Teams' AI features (called Copilot in Teams) have faced adoption challenges — the same ones affecting Microsoft's overall Copilot story we covered earlier this week.
Slack's aggressive AI feature push creates real competitive pressure. Indian IT decision-makers who've been defaulting to Teams because it comes bundled with Office subscriptions might now have reason to reconsider, especially for teams that don't need the full Microsoft 365 stack.
TamilTech's take
The reusable AI skills concept is the smartest thing in this announcement. Most AI features are still in the mode of "ask the AI and it responds" — which is useful but still manual. Reusable skills flip this: you define the workflow once, and the AI executes it on demand with minimal friction. That's closer to actual workflow automation than most AI chat features. The desktop monitoring is the one to watch carefully — the potential for productivity benefit is real, but so is the potential for invasive workplace surveillance depending on how companies implement it. Slackbot was always capable but underused because it wasn't good enough to trust with real work. These features change that calculus. Whether Indian enterprises will adopt at speed depends on how quickly the paid plan features roll out and whether the privacy controls are actually user-respecting rather than just checkbox features in the terms of service.
Comments (0)
Be the first to comment!