Slack has introduced new agentic automation capabilities that bring its Workflow Builder, Slackbot and third-party AI agents together in a single workspace. The update allows businesses to create automated processes that can trigger AI agents, delegate research and writing tasks, and return generated results without employees having to switch between applications or manually copy information.
Announced on November 18, 2026, the update positions Slack as an automation hub for businesses adopting AI agents. Companies can now use workflow events to start conversations with Slackbot or other agents, schedule recurring agent tasks and introduce human approval steps before AI-generated information is sent to business applications. Slack says the capabilities are designed to make AI agents part of everyday business processes rather than isolated chatbots.
Key Takeaways
- Slack is bringing Workflow Builder, Slackbot and third-party AI agents together.
- New workflow steps allow businesses to send tasks directly to Slackbot or external agents.
- Slackbot Tasks can start automatically when events occur in Slack.
- Agent Tasks allow users to schedule recurring prompts for third-party agents.
- AI-generated content can be reviewed and edited by employees before reaching business applications.
- The update supports workflows involving tools such as Jira, Google Sheets and Salesforce.
- Workflow Builder is included in Slack Pro, Business+ and Enterprise plans.
Slack Turns Workflow Builder Into an AI Automation Hub
Traditional workflow automation systems are designed to perform predictable tasks.
For example, a company can configure a workflow to send a welcome message whenever a new employee joins, notify a manager when a customer submits a form or update a project channel when a task changes status.
However, these systems often struggle when a process requires reasoning, research or content generation.
An employee may need to leave the workflow, open an AI assistant, provide the necessary information, copy the generated response and return it to the original application.
Slack wants to eliminate this manual process.
Its new approach connects Workflow Builder directly with Slackbot and third-party agents. This allows a workflow to hand a task to an AI system and bring the result back into the same process.
For example, when a new employee joins a company, a workflow could collect their details, send the information to an AI agent and automatically generate a personalised welcome message.
The employee’s manager could then review the draft before it is sent.
This makes the AI assistant part of the workflow rather than a separate tool employees must operate manually.
Employees Can Add Slackbot and AI Agents to Workflows
One of the main additions is the ability to start conversations with Slackbot or an external agent directly from Workflow Builder.
Slack has introduced two new workflow steps:
1. Start Conversation With Slackbot
This step allows a workflow to send a predefined prompt to Slackbot when the process reaches a particular stage.
Developers and workflow administrators can include workflow variables, skills and knowledge sources to provide additional context.
When the workflow runs, Slackbot opens in the user’s sidebar with the prompt already submitted.
For example, a daily stand-up workflow could ask Slackbot to prepare an employee’s update based on their previous day’s work.
The employee can review the generated response and share it with the team.
This removes the need to manually write the same prompt every morning.
2. Start Conversation With an Agent
The second step allows businesses to connect third-party or custom AI agents to their workflows.
Administrators can select an available agent, configure a prompt and add workflow variables.
When the process runs, Slack starts a conversation with that agent and provides the relevant information.
Consider an IT support workflow.
An employee reports that their VPN connection keeps disconnecting. The workflow can initially provide troubleshooting instructions. If the problem continues, the employee can select an option to involve the IT help desk agent.
The agent receives the original question and relevant context, allowing it to investigate the issue without requiring the employee to explain everything again.
Slackbot Tasks Can Automate Work Based on Real Events
Another major feature is Slackbot Tasks.
These tasks allow Slackbot to begin working automatically when specific events occur inside Slack.
Instead of waiting for an employee to open the AI assistant and submit a prompt, the system can respond to an event such as:
- A new message appearing in a channel.
- Someone adding a reaction to a message.
- A new item being added to a Slack List.
- An existing list item being updated.
For example, a customer support team could configure Slackbot to prepare a suggested response whenever an employee reacts to a customer message.
The system could summarise the issue, draft a reply and make the result available for review.
Another example involves project management. When a new item is added to a Slack List, Slackbot could prepare a status update using the available information.
These capabilities are intended to reduce repetitive administrative work and help teams respond more quickly to incoming information.
Slack says integrations with calendar, Salesforce and service tools such as PagerDuty are planned as additional triggers.
Agent Tasks Allow Businesses to Schedule Recurring AI Work
Slack has also introduced Agent Tasks, which allow users to save prompts for third-party agents and run them automatically on a schedule or in response to Slack events.
This addresses a common problem with AI assistants: employees often have to remember to repeat the same instructions.
For example, a sales manager may need a weekly report showing the largest opportunities in the company’s pipeline.
Previously, the manager might have opened a sales agent every Wednesday, submitted the same prompt and waited for the results.
With Agent Tasks, the prompt can be saved and scheduled.
Slack’s example involves a sales agent running every Wednesday at 8:45 a.m. to identify opportunities owned by a particular employee, rank them by deal size and suggest next steps.
The agent then posts its results in a conversation thread, where the employee can ask follow-up questions.
This could make recurring AI-assisted activities easier to manage across sales, marketing, operations and customer support teams.
Human Approval Remains Part of the Automation Process
One of the most important aspects of Slack’s announcement is its emphasis on human review.
Businesses may be reluctant to allow AI systems to directly modify important records or send information to customers without supervision.
A mistake in a spreadsheet, customer database or project-management ticket could create additional problems.
To address this concern, Slack automatically adds a review step after an AI step in Workflow Builder.
When the workflow reaches that stage, it pauses and displays the generated content to the appropriate employee.
The employee can edit the information and approve it before the workflow continues.
For example, a company could create an automated process that prepares a weekly team status report using recent activity.
Slackbot generates the report, but a manager reviews the content before it is added to Google Sheets.
The same approach can be used for incident reports that need to be reviewed before being added to Jira.
This design allows companies to automate repetitive work while keeping employees involved in important decisions.
How Slack’s AI Automation Could Help Businesses
The new capabilities could affect several business functions.
| Business function | Possible AI automation |
|---|---|
| Human resources | Generate personalised employee onboarding messages |
| Sales | Prepare recurring pipeline reports and suggest next steps |
| Customer support | Summarise incoming complaints and draft replies |
| IT support | Investigate technical issues and recommend solutions |
| Project management | Generate status reports and update project records |
| Marketing | Draft campaign updates and prepare content summaries |
| Operations | Route requests and prepare reports based on workflow events |
The main advantage is that businesses can combine predictable workflow steps with AI-powered reasoning.
A traditional workflow can determine when something should happen, while an AI agent can handle tasks that require interpreting information or generating a response.
This combination could make business automation more flexible.
However, the actual benefits will depend on the quality of the connected agents, the accuracy of their results and how well companies configure their workflows.
Slack Faces Competition in the AI Agent Market
Slack’s announcement comes as enterprise software companies compete to become the primary interface for AI-powered work.
Microsoft has been expanding Copilot and its agent capabilities across Microsoft 365. Salesforce is developing Agentforce to automate customer relationship management and other enterprise processes.
Google is also expanding AI capabilities across its Workspace ecosystem.
Slack’s strategy is to bring employees, business applications and AI agents together inside the same communication platform.
Its advantage is that teams already use Slack to discuss projects, share information and coordinate work.
By integrating agents into those conversations and workflows, the company wants to reduce the need to move between separate AI applications.
Slack has also been developing its broader agent ecosystem. Its earlier announcements included the Slack Agent Kit, agent discovery capabilities and integrations with external AI platforms.
The new Workflow Builder features extend that strategy by connecting agents to actual business processes.
What This Means for AI Agent Developers
The announcement is also relevant to developers building custom AI agents.
Previously, developers often needed to build separate interfaces and automation systems to connect their agents to business applications.
Slack’s approach gives available agents a way to participate in existing workflows.
Businesses can select an agent, configure a prompt and use it as part of an automated process.
This could make it easier for developers to distribute specialised agents to companies already using Slack.
For example, a developer could create an agent focused on analysing financial reports, reviewing customer support tickets or helping employees troubleshoot IT problems.
If that agent is available within a company’s Slack workspace, employees can access it through conversations and workflows rather than switching to another application.
Slack’s existing agent ecosystem and developer tools could therefore become increasingly important as businesses look for practical ways to deploy AI systems.
Availability and Pricing
Slack says Workflow Builder is included in its Pro, Business+ and Enterprise plans.
The company announced that businesses can add Slackbot or available third-party agents to workflows using the new capabilities.
However, access to individual agents and integrations can depend on workspace configuration, administrator permissions, licensing and rollout availability.
Businesses should also check the requirements of external applications before deploying automated processes that access customer information or modify business records.
Slack has not announced a separate standalone price for the new automation capabilities in its November 18 announcement.
The Bigger Picture
Slack’s latest update reflects an important evolution in enterprise AI. Businesses are moving beyond using chatbots simply to answer questions and are increasingly looking for agents that can participate in actual work processes. Combining workflow automation with AI reasoning could allow companies to automate tasks that previously required employees to move information between different applications.
The key challenge will be ensuring that these systems remain reliable, secure and properly governed. AI agents can generate incorrect information or misunderstand instructions, particularly when dealing with complex business data. Slack’s built-in review steps are therefore an important part of its strategy. The long-term value of agentic automation will depend not only on how much work AI can perform but also on whether businesses can trust the results.
Looking Ahead
Slack is likely to continue expanding its agent ecosystem and the number of business events that can trigger automated tasks. Planned integrations with calendar, Salesforce and service platforms could make the system more useful for companies operating across multiple applications. The adoption of these features will also depend on how easily administrators can manage permissions, monitor agent activity and connect existing business systems.
For businesses, the announcement signals that AI agents are becoming part of mainstream workplace automation rather than remaining separate experimental tools. If Slack can make it simple to combine automated processes, AI-generated decisions and human approval, it could strengthen its position as an important platform for enterprise AI. The next test will be whether companies can achieve measurable productivity improvements without introducing new operational and security risks.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.



