Salesforce, Inc. on the 12th unveiled a feature that integrates Anthropic's artificial intelligence (AI) assistant "Claude" bidirectionally with its collaboration tool "Slack" and its enterprise work platform "Agentforce 360."
The core of this feature is that AI goes beyond simply searching for and organizing information to grasping a company's workflow and supporting task execution. For example, a user can have Claude organize meeting agendas or plans to create a draft document, then share it to Slack to continue follow-up work. Going forward, features will be added to automatically execute tasks such as customer management, service handling, and approvals based on Claude's analysis.
This bidirectional integration is built on Anthropic's "Model Context Protocol." Salesforce, Inc. said it is an open standard technology that safely connects AI to external systems, designed so that when Claude is linked with Slack or Agentforce 360, AI can operate within existing security environments without any direct data movement.
Accordingly, it can be used safely even in sectors where data sensitivity is high and security is critical, including finance, healthcare, and the public sector.
Park Se-jin, head of Salesforce Korea, said, "Salesforce, Inc. is focusing on supporting a shift to an 'agentic enterprise,' where AI not only serves as an assistant but actually executes business tasks and coexists as a digital colleague within the organization," adding, "This bidirectional expansion with Claude will play a key role in driving practical work innovation and business innovation based on organizational data, founded on the organic integration with Slack and Agentforce 360."