Snowflake CoWork Evolves into a Personal AI Agent for Knowledge Workers, Driving the Agentic Enterprise

Snowflake (NYSE: SNOW), the AI Data Cloud company, today announced new AI innovations at Snowflake Summit 26 that further expand AI adoption and operations across the enterprise. The latest capabilities of Snowflake CoWork (formerly Snowflake Intelligence) enable organizations to move more quickly from insight to action. Built for knowledge workers, CoWork is a personal AI agent that supports smarter ways of working. The latest enhancements to CoWork aim to provide more advanced AI capabilities to knowledge workers who struggle to maintain context across dispersed data, tools, and workflows. With new features such as Artifacts, Cortex Sense, and personalization capabilities, CoWork integrates data, context, and AI into a single agent experience, enabling business teams to make decisions and execute tasks while leveraging AI in a secure and trusted environment. Simultaneously, for data teams, the new 'Cortex Training' feature makes it easier to extract insights from data and train models at scale. Currently, over 13,900 customers worldwide trust their critical data to Snowflake. Snowflake applies AI directly to that data within a secure and governed environment. Snowflake also offers customers the choice and flexibility to select the best AI models for their business needs. Through Snowflake Cortex AI, its suite of AI products, customers have access to leading AI models from Anthropic, OpenAI, Google, Meta, Mistral, DeepSeek, and SpaceXAI, announced today. As a result, Snowflake's AI not only delivers high performance but also provides a trusted foundation at enterprise scale, helping organizations achieve practical business outcomes and sustainable competitive advantage. Christian Kleinerman, Senior Vice President of Product at Snowflake, stated: "Companies with a true competitive advantage are those that empower every employee to harness the full power of AI. CoWork provides every employee with a personal AI agent that can work across their enterprise data and tools. This allows them to gain insights, take action, and collaborate in a unified, trusted, and governed environment. Our vision is a future where AI acts not just as a provider of insights, but as a 'control plane' that connects data, context, and enterprise systems to drive action across the organization." To deliver a highly personalized AI experience, CoWork builds a common understanding of enterprise data, linking organizational business context with each user's role to provide optimized insights and actions for each individual. This allows users to leverage their skills to transform daily tasks into automated workflows and directly connect with business tools like Google Drive, Salesforce, and Slack through Model Context Protocol (MCP) connectors. As a result, teams can move ideas into action quickly across the various systems they use. These MCP capabilities are built upon Snowflake's recently announced plan to acquire Natoma. Natoma's enterprise MCP platform provides secure connectivity, governance, identity-aware authorization, and auditability for AI agents operating across enterprise systems and tools, supporting safe and controlled AI use in corporate environments. For complex queries requiring deeper analysis, Deep Research performs multi-step reasoning and analysis across an organization's structured and unstructured data to conduct comprehensive investigations. Unlike general consumer AI assistants or research tools, Deep Research operates within the enterprise's governed data environment, analyzing documents and business data to provide actionable insights with cited sources. This enables users to understand not just the 'what' but the crucial 'why' behind the numbers. Built by Snowflake's AI Research Team, an agent swarm orchestration system enables Deep Research to achieve state-of-the-art performance across structured and unstructured enterprise data. On Snowflake's Hybrid Deep Research Benchmark, it demonstrates over 30% higher performance compared to single-agent systems. Thousands of Snowflake users are already leveraging CoWork to improve operational efficiency and accelerate decision-making. Synopsys, Whoop, and Under Armour are among these examples. With this adoption, CoWork is evolving beyond a passive question-and-answer AI to an entity that understands context and supports decision-making.