Salesforce has launched its Enterprise AI Harness, a composable architecture designed to unify six core capabilities—context, agency, action, governance, security, and models—into a single trusted foundation for enterprise AI. Built on the company’s existing Data 360, Informatica, MuleSoft, Tableau, Agent Fabric, Agentforce, Salesforce Guardian, and Platform technologies, the Harness enables enterprises to leverage proprietary context—customer relationships, workflows, semantics, and history—as the basis for AI reasoning and action.(salesforce.com)
At the heart of the offering is a new AI Control Plane, which provides centralized discovery, registration, identity and policy enforcement, lifecycle management, performance evaluation, behavioral observation, and cost control for AI agents and models—whether native to Salesforce or third-party. This gives organizations a unified layer of visibility and governance as AI agents proliferate across teams and systems.(salesforce.com)
Salesforce emphasizes that enterprises can adopt the full Harness or selectively integrate individual capabilities alongside their existing technology stack. The architecture is designed for openness and composability, allowing customers to mix Salesforce components with third-party models, agents, and systems.(salesforce.com)
This launch marks a significant step in the evolution of enterprise AI, addressing the growing need for scalable, secure, and governed agentic systems. By embedding AI within trusted enterprise context and providing a unified control plane, Salesforce aims to help organizations move beyond isolated pilots toward operationalized, enterprise-wide AI deployments.