News

After a long day of meetings, you find hundreds of messages waiting for you. Is there a brewing crisis, an important announcement or just an overzealous company potluck chat group? It would take you about 30 minutes to sieve through them, but you just don’t have the time.  What if there was a way to cut through the clutter in mere seconds?  Slack believes it has found an answer. Since February, the AI-powered work platform has rolled out , its unique take on harnessing generative artificial intelligence (AI) to help businesses find information quickly and without hassle.  “Being able to pick the day, week or month I’d like to catch up on with Slack AI has been so impactful,” says Mr Jason Morita, product owner at ProService Hawaii, a human resource service provider. “Conversation summaries save the day after meeting marathons: I use them to stay informed on what I missed while I take care of something else.”  Mr Morita is just one of the few early users of Slack AI who have already benefited from the tool’s intuitive generative AI features, which help simplify how AI is used in the workplace. Sharing its vision of AI for businesses, Slack’s vice president of product management, Ms Jackie Rocca, explains: “For the past 10 years, Slack has become a hub for our clients’ projects, data and information. So there’s this incredible collection of information, but sometimes it can be challenging to pull out the most important nuggets of information that matter.  “We designed Slack AI to help our customers unlock the most value out of their knowledge and find information quickly.”  With Slack AI, users can now generate summaries of chat threads and entire channels, so employees can get up to speed quickly and without lengthy meetings. This can save them an average of 97 minutes per week, Slack found. Another new feature is the integration of Slack AI into search queries. Through keyword-based or natural language queries, Slack AI can provide personalised and contextual results.  Early Slack AI users have reported positive experiences with the Slack AI-powered search function, shares Ms Rocca. For instance, in a virtual meeting, a sales representative had to answer her customer's question about a compliance-related feature. The normal procedure would involve asking the customer to wait while she checked with the subject-matter experts on the feature. However, with the Slack-AI-powered search, the sales representative could find the answer in a channel discussion within seconds, allowing her to continue the conversation with the customer immediately.  “This is an example of a real business benefit that Slack AI achieves for our users,” says Ms Rocca.  Slack AI can also help employees stay in the loop of other work developments more efficiently. In April, Slack AI welcomed a new recap feature, a newsletter-like knowledge management tool which automatically summarises selected channels.  This “digital morning digest”, as Ms Rocca calls it, is useful for managers or adjacent teams who want to keep track of a project, but not necessarily the granular details.  Recap summarises the most important developments daily, so subscribers do not need to trawl through long threads of individual messages. This way, users can efficiently keep abreast of multiple developing projects without feeling overwhelmed by the fine print.  Even with the advent of AI being integrated into many businesses, human control remains the focus of Slack’s feature design philosophy.  In a , Ms Rocca says: “We are focused first on really keeping the human in the loop and making sure that you are always in control.”  This means that Slack AI is constructed in a way that builds trust and emphasises human control over both the data accessed by the AI and the actions the AI takes.  Slack AI combines the capabilities of AI with the strength of human judgment to make AI more effective and trustworthy. All the answers provided by Slack AI are drawn from existing company information, not external sources, and all responses are accompanied by citations, allowing users to check the context of the provided information.  This way, users can be assured that the information they are accessing is accurate while saving time to focus on more meaningful, valuable work. In addition, Slack users play an important role in improving Slack AI. With every generated answer or summary, users can rate the response and provide feedback to the development team.  Slack AI also follows the same enterprise-grade data security and compliance standards as the Slack app. Slack AI’s self-hosted model means customer data is not used to serve other clients or to train AI models.  As AI becomes more common in the workplace, Slack is helping its users tap into its potential as an accessible technology for all. “There are some AI tools where you almost need to be a prompt engineer to get the most out of them,” remarks Ms Rocca.  “We designed Slack AI for the everyday person learning how to make use of AI in the workplace. Our goal was to make it accessible and easy to use in your daily workflow.” For instance, search queries are automatically enhanced with AI, so there is nothing extra the user has to do apart from simply keying in the query in the pre-existing search box. Thread and channel summaries are also accessible with one click on the Slack AI icon within the app. Users can then customise their search range from the last seven days or set a custom date range.  Slack’s native AI builds on a large collection of over 13,000 third-party AI-powered apps already compatible with the Slack ecosystem. Ms Rocca says: "We want to be inclusive of all the other tools and ensure you can bring those into Slack as well." These include the writing tool Notion and conversational chatbot search engine Perplexity, which can help automate repetitive tasks. Slack is also building a native AI integration with .