Salesforce bets enterprise AI edge lies in context and control

Insider Spotlight 

• Salesforce argues proprietary business context—not the underlying AI model—will increasingly determine how companies differentiate with agentic AI
• Its Enterprise AI Harness combines six capabilities covering context, agency, action, governance, security and models
• A new AI Control Plane is intended to give enterprises centralized visibility and control over Salesforce and third-party agents and AI
  
  
  
  

As artificial intelligence (AI) models become more widely available, Salesforce is betting that enterprises will gain a more durable advantage from the proprietary customer data, business processes and controls they can securely connect to AI. 

The company says this context can help agents move beyond generating answers to taking predictable actions across business systems.

Enterprise AI Harness 

Salesforce on Sept. 11 introduced its Enterprise AI Harness, a composable architecture designed to give agents shared business context while applying governance and security controls. It draws on technologies including Data 360, Informatica, MuleSoft, Agent Fabric, Tableau, Agentforce, Salesforce Guardian and the Salesforce Platform.

Architecture groups 

Enterprise questions often require information scattered across CRM, ERP, contracts, analytics and previous interactions. Salesforce says connecting systems alone is insufficient because AI must interpret that information for a particular customer, decide what action to take and operate within established policies.

The architecture groups those requirements into Trusted Context, Trusted Agency, Trusted Action, Trusted Governance, Trusted Security and Trusted Models. Companies can use the capabilities together or selectively alongside existing technology, third-party models, agents and systems.

“The Agentic Enterprise won’t be defined by which model a company chooses. Models will keep changing, and intelligence will increasingly be available everywhere. What will differentiate an enterprise is the trusted, proprietary context it brings to that intelligence—starting with the customer—and its ability to turn that context into action securely,” Salesforce president and chief platform and engineering officer Rohan Kumar said in a press statement.

AI control plane

Salesforce is also introducing an AI Control Plane for discovering and registering AI agents, establishing identity and policy, managing lifecycles, evaluating performance, observing outcomes and controlling costs across Salesforce and third-party AI.

Unified experience

Many underlying technologies are already available, while new capabilities and a unified experience are planned to begin rolling out in early fiscal FY28. Salesforce said further availability, packaging, pricing and upgrade details will come closer to general availability. —Vanessa Hidalgo | Ed: Corrie S. Narisma