Salesforce has introduced the Enterprise AI Harness, an architecture designed to give AI agents a shared understanding of a business’s data, policies and controls so they can act reliably without being rebuilt for every use case.

The Harness brings together six capabilities spanning context, agency, action, governance, security and models, drawing on existing technologies including Data 360, Informatica, MuleSoft, Agent Fabric, Tableau, Agentforce and Salesforce Guardian. Alongside it, a new AI Control Plane gives businesses a single place to discover, monitor and manage agents and AI activity across Salesforce and third-party systems.

Built to work across MCP, APIs and other AI experiences

Salesforce says the Harness is built headlessly, with capabilities accessible through MCP, APIs, Skills and plug-ins, extending its reach beyond Salesforce applications into AI experiences including Claude, Slack, Microsoft Teams and Agentforce. Customers can use all six capabilities as one system or adopt individual pieces alongside their existing technology stack.

“The Agentic Enterprise won’t be defined by which model a company chooses. Models will continue to change, and intelligence will increasingly be available everywhere. What will differentiate an enterprise is the trusted, proprietary context it brings to that intelligence,” said Rohan Kumar, President, Chief Platform and Engineering Officer at Salesforce.

Many of the underlying technologies are already available, with the unified experience and new capabilities rolling out from early fiscal FY28. Existing Salesforce customers will be able to upgrade eligible investments to access new capabilities as they launch, with pricing and packaging details to follow closer to general availability.