24 - 08 - 2026

Slack Turns AI Coding Into a Team Sport with New ‘Slack Code’

Slack has introduced Slack Code, a new AI-native coding experience designed to move software development from private developer sessions into a shared, collaborative environment where humans and AI agents can build together.

The new capability introduces dedicated code channels, giving teams a workspace specifically designed for agent-driven software development. Instead of coding agents operating in isolated terminals or browser tabs, teams can now follow their work, review changes, provide feedback and approve deployments within Slack.

The move comes as coding agents from companies including Anthropic, Cognition, GitHub, OpenAI and Vercel increasingly automate software development tasks. Slack is positioning Code as the next evolution of that trend—from individual AI-assisted coding to multiplayer development involving entire teams.

“AI only creates value when it’s part of how a team actually works, not something people go do alone in another tab. Code channels bring that work into the open, where the rest of the team can see it, shape it, and build on it,” said Rob Seaman, EVP and GM of Slack, Salesforce.

From conversation to working code

Each Slack Code channel is linked to a specific project, keeping agent-driven development separate from ordinary conversations. Teams can tag a coding agent from an existing Slack conversation, triggering the creation of a dedicated code channel for the task.

Within the channel, participants can follow the agent’s conversation and plan, inspect code diffs, view live previews and provide instructions or corrections. Once the task is complete, the channel can automatically archive while preserving its history as a searchable record.

The approach is designed to make software development accessible beyond engineering teams. A product manager, for example, could identify a bug in a Slack conversation, ask an AI agent to investigate and propose a solution, and then bring an engineer into the code channel to review the changes before the agent opens a pull request.

Slack argues that this model could reduce the handoffs traditionally required between product teams and developers while keeping human oversight in critical stages of development.

Slack Code is being launched with an ecosystem of coding-agent partners including Anthropic’s Claude, Cognition’s Devin, GitHub Copilot, OpenAI’s ChatGPT and Vercel’s agents.

“So much engineering work already starts as a conversation in Slack with Claude Tag: a bug report, a piece of feedback, a ‘what do you think about…’ Slack Code gives teams a dedicated channel where the people, the context, and Claude work together from the start,” said Cat Wu, Head of Claude Product at Anthropic.

Vercel CTO Malte Ubl said code channels give agent-based development a dedicated environment where teams can collectively steer an AI agent while keeping context and history in one place.

GitHub Chief Product Officer Mario Rodriguez described the integration as part of GitHub’s broader approach in which humans set direction while agents help close the development loop.

Slack Code is designed around visibility rather than autonomous development happening outside the team’s view. Members can see proposed changes, inspect diffs and run live HTML previews before software is shipped.

For high-stakes actions such as production deployments, the system keeps a human in the approval loop, allowing an expert to review and sign off on the agent’s work.

Agents also inherit Slack’s existing permissions, security model and administrative controls, according to the company, allowing organisations to manage AI agents within their existing Slack environment.

Cognition President of New Enterprise Jeff Wang said agents must operate like effective teammates, contributing when appropriate while avoiding unnecessary disruption. Slack sees coding as the starting point rather than the endpoint for its new agent-based collaboration model.

The company plans to open its code-channel APIs more broadly so organisations can bring custom agents into shared workflows beyond software development. That could eventually include teams using agents for activities such as marketing campaigns, document reviews and other specialised workflows.

Slack is also expanding its broader agent experience with an Agents tab, agent direct messages and its “Add to Slack” functionality, which allows organisations to bring agents built with platforms such as Lovable, Vercel, OpenAI, LangChain and others into Slack.

According to Rebecca Wettemann of Valoir Research, the significance extends beyond individual productivity.

“For companies to unlock the full value from AI, it has to move beyond individual productivity gains to operational transformation. Slack Code enables AI to move beyond individual copilots to embedded, contextual AI orchestration to deliver real AI ROI at scale,” she said.

Slack Code is available across Slack plans for the current release, with access to the respective partner AI agents required. The company said pricing and packaging may change and that availability can vary by region and customer agreement.

With Slack Code, the company is effectively betting that the next phase of AI-assisted software development will not be human versus machine or developer versus agent, but teams of humans and AI agents working in the same digital workspace—with shared context, visible progress and human judgment remaining part of the development loop.

By Subramanya Joshi