Slackbot becomes the new AI agent within Slack

Slackbot, Slack's veteran automated assistant, takes a major turn and becomes a job-focused artificial intelligence agentSalesforce, Slack's parent company, wants this new Slackbot to be the natural entry point to AI in the company, leveraging the data that already circulates daily through channels, direct messages, and connected applications.

The company presents Slack as a kind of conversational operating system for the officewhere people and AI agents coexist in the same channels. In that context, Slackbot becomes a team partner that understands conversation history, business context, and each user's permissions., forks capable of answering questions, generating content, and executing actions without having to leave the application.

Slackbot as a personal work agent

The new version of Slackbot is designed as a a highly personal assistant for each employee, natively integrated into all supported Slack workspaces. There's nothing to install, no need to activate extra plugins, and no need to learn new commands.Simply start a conversation with Slackbot, just as you would with any colleague.

Salesforce emphasizes that There is no need to train models or design complex promptsSlackbot “starts” directly with the existing context within the organization: messages, files, channels, user profiles, and connected applications. From there, it can interpret questions like “What did we decide about the Q4 budget?” and return a selected response, linked to specific conversations and documents, instead of a simple generic search.

Parker Harris, co-founder of Salesforce and chief technology officer of Slack, defines this new Slackbot as “an agent, a super-agent who acts as your assistant”Internally, Slack has been using it for months before the general launch, and according to the company itself, it has become the most widely adopted AI tool by its own employees.

Beyond offering texts, the agent can adapt to each person's communication styleto adapt to the team's style and learn their usage preferences. The goal is for interacting with Slackbot to feel more like talking to a colleague than navigating a technical interface.

What can the new Slackbot actually do?

Salesforce's strategy isn't limited to answering specific questions: the new Slackbot is designed to reduce manual work and orchestrate complete workflows within the workspace. Among its main capabilities, the company highlights several functions that are especially relevant for European companies that already work in hybrid or remote environments.

On the one hand, it can act as contextual search engineInstead of simply locating a message, Slackbot understands which project or client the question refers to and locates the most relevant information across channels., threads, documents, and attachmentsas well as in connected corporate data. This allows, for example, asking someone to "find the file Marta sent me with the fourth-quarter results" and obtaining the correct document without having to search through hundreds of messages.

On the other hand, the agent serves as productivity tool for repetitive tasksYou can draft emails, prepare meeting notes and summaries, generate reports from internal conversations, or create task lists from a discussion thread. All of this is done without leaving Slack.This makes it easier for teams from Spain or the rest of Europe that already use the platform to further centralize their day-to-day operations.

Slackbot is also capable of schedule meetings, find gaps in the calendar and set reminders for important deadlines, provided you have permission to access the relevant information. The idea is that the user can chain several actions in a single conversationinstead of jumping between different tools.

Connection with Salesforce CRM and business data

One of the cornerstones of this evolution is the close integration with Salesforce. Thanks to the database of Salesforce CRM Slackbot can already combine the ecosystem's real-time APIs conversational data with business metrics safely.

In sales scenarios, for example, an agent can gather key information before a meeting, such as a customer's interaction history, ARR (annual recurring revenue), or open issues, and cross-reference it with conversations the team has had across sales channels. With this raw data, they are able to directly generate a briefing or a strategic report to prepare for a call or a presentation.

Similarly, teams can talk to Slackbot in natural language and ask it to Prepare pipeline summaries, business proposals, or account updatesusing real-time data from Salesforce. The company's intention is for Slack to evolve from just a corporate chat platform into a command center of the “agentic company”where AI takes an active role in performing the work.

For those developing solutions on the platform, Salesforce has made new tools available, such as APIs focused on real-time conversational data, which can be used to create custom agents or advanced integrations that rely directly on the Slackbot context.

Open architecture and collaboration with other AI models

Another key aspect of the approach is that Slackbot is not limited to Salesforce's own technology. Under a open architecture strategyThe agent can work with external models such as those from OpenAI (ChatGPT) or Anthropic (Claude), as well as with Agentforce, Salesforce's AI agent product.

In practice, this means that Slackbot can behave like a multi-agent coordination centerFrom a single conversation, the user could trigger actions involving different specialized AI models, while Slackbot orchestrates the workflow and delivers a consistent result. All of this is done under Salesforce's trust layer, whose mission is to... prevent leaks and protect sensitive data.

Looking to the medium term, Slack anticipates that Slackbot will extend beyond text-based functionality. Among the plans Parker Harris has hinted at is the addition of... voice and internet browsing capabilitiesso that the agent can accompany the user beyond written channels and offer information or perform actions in more dynamic contexts.

Furthermore, the company suggests that Slackbot could be the natural way to interact with both Agentforce and other third-party agents, acting as unified conversational interfaceThe strategy aligns with trends identified in recent studies on AI productivity, which indicate that internal agents are often particularly effective in back-office processes and internal support.

Privacy, security and control in the business environment

Since it's a tool that accesses conversations, files, and business data, one of Salesforce's recurring messages is that... “enterprise-level confidence”According to the company, Slackbot can only see information that the user already has access to, always respecting the permissions configured in workspaces, channels, and applications.

This means that the agent is informed by messages and filesHowever, it is limited by the same access controls that apply to the rest of the platform. Furthermore, Salesforce assures that conversations with Slackbot are not restricted. They are not used to train AI modelsThis is especially relevant for European organizations subject to strict data protection regulations.

From an internal governance perspective, Slack allows administrators can disable or restrict Slackbot at the organization level if they deem it appropriate. Regarding access to conversations with the agent, the company indicates that administrators do not, by default, have additional visibility into these exchanges beyond the standard auditing capabilities already present on the platform.

With this approach, Salesforce seeks to balance the potential of AI in the workplace with the security and compliance requirements of companies, both in Spain and in the rest of Europe, where the management of personal and corporate data It is highly regulated.

Availability, plans and global deployment

The new Slackbot is now available generally available for Slack business customers, starting with the Business+ and Enterprise+ plans. According to official announcements, the rollout is being done gradually and is expected to reach users from all over the world, including the European market, in the next activation phases.

For organizations that already use Slack in their daily operations, there is no need to incorporate additional products or manage complex installations: Slackbot arrives activated as part of the platform itself.However, the decision to keep it in production or not rests with the administrators of each workspace.

In parallel, Salesforce continues to advance related product releases, such as feature packages included in new release waves, where Slackbot and developer tools feature prominently. The goal is that, over time, More and more corporate systems are integrating with the agentso that AI has visibility over a wider range of data without leaving the company's security framework.

The company admits that this is just the beginning of a new stage for Slackbot as an AI agent. From now on, its evolution will depend on both technical improvements and... actual adoption by organizationswhich will determine which use cases become established and which ones are relegated to the background.

With all these changes, Slackbot is no longer that simple bot that responded to basic commands—like the answering machines in Slack— to become a central piece of Salesforce's AI strategy: a An agent integrated into Slack that leverages conversational context, business data, and an open architecture to help European companies work more efficiently, automate processes, and concentrate corporate intelligence in a single conversational access point.