Slack Adds AI Features to Its Messaging Platform

Business messaging app Slack is rolling out Slack AI, a set of generative AI features, including AI-assisted search within conversations, channel recaps, and thread summaries.

Slack’s generative AI experience is a paid add-on for enterprise plans. It’s only available in English across the US and the UK, with more languages and plan models coming soon.

The new AI features harness the power of conversation histories to find relevant information in just a few clicks. Enterprise-level businesses like Anthropic and SpotOn trialing the Slack AI pilot program report that it has saved them an average of 97 minutes per user per week by “using Slack AI to find answers, distill knowledge and spark ideas.”

One of the new Slack AI features enables a faster search through conversations, eliminating the need to sift through results manually to find relevant information. Users can now search with conversational questions and get personalized answers with direct citations to relevant Slack messages from public and personal communications. If users are lacking context, Slack AI can provide suggestions.

The new AI-powered search experience can help users learn about new projects, review company policies, and get insights about past decisions significantly faster than before.

Furthermore, channel recaps can generate highlights within any channel, saving users from reading piles of messages. Users can personalize their channel recaps asking for information from unread messages or requesting summaries from a specific date range.

While recaps can summarize key information in a busy channel, thread summaries can do the same for long conversations. Each summary also contains sources pointing to the relevant messages.

According to Slack, channel recaps and thread summaries “help you cut through the noise and amplify what’s most important.” This can be beneficial in multiple scenarios, from giving new employees more context about the company and their roles to catching up after a long break from work.

Slack assured users it won’t share customer data with third parties for AI development purposes or use the data to train proprietary large language models (LLMs). The mix of LLMs powering Slack’s generative AI experience is hosted on its infrastructure, meaning data remains in-house.

Additionally, the business messaging platform announced “many exciting features on the horizon.” One is letting users create digests from channel highlights about things that don’t require immediate attention. Slack is also developing an AI integration for Einstein Copilot, the new conversational AI assistant for Salesforce CRM.