Salesforce has announced the Enterprise AI Harness, a new architectural foundation designed to provide artificial intelligence agents with a shared understanding of enterprise operations while enforcing strict security, governance, and business controls.

As corporate workflows shift toward the “Agentic Enterprise”—where autonomous digital agents handle complex tasks such as analyzing events, making operational decisions, and executing cross-system actions—businesses face growing challenges in operating these tools reliably and securely across disparate departments.

The Enterprise AI Harness addresses this by consolidating the underlying infrastructure required for enterprise AI into a single composable system, eliminating the need for organizations to build separate governance and context layers for every individual agent.

Bridging Open-Ended AI with Business Rules

Answering routine operational questions—such as whether a customer order can be fulfilled immediately—typically requires pulling context from multiple isolated systems, including CRM platforms, ERP inventory tools, contract databases, and analytics engines.

Salesforce’s new architecture connects AI reasoning with deterministic business rules, allowing agents to understand live customer context, evaluate entitlements, check inventory, and execute actions predictably within set policy boundaries.

Every execution and interaction within the system feeds into a compounding intelligence loop, continually updating the organization’s proprietary memory and operational context over time.

“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,” said Rohan Kumar, President and Chief Platform and Engineering Officer at Salesforce.

“That’s what we’re building with Salesforce’s Enterprise AI Harness: a system where every interaction and action can create new context and intelligence, making the enterprise’s context richer over time while operating with the security, governance, and control businesses need to trust AI at scale.”

The Six Core Capabilities

The architecture brings together technologies from Data 360, Informatica, MuleSoft, Agent Fabric, Tableau, Agentforce, Salesforce Guardian, and the Salesforce Platform. It is structured around six unified capabilities:

• Trusted Context: Combines real-time data, metadata, semantics, customer history, and enterprise rules to provide AI with a unified view of the business.
• Trusted Agency: Supplies agents with reasoning, planning, state memory, and orchestration while enforcing fixed business rules where certainty is mandatory.
• Trusted Action: Connects AI directly to APIs, workflows, applications, and tools, allowing agents to execute tasks like reserving inventory or updating orders.
• Trusted Governance: Manages data lineage, policy enforcement, quality controls, and operational guardrails across all AI activity.
• Trusted Security: Enforces identity management, user permissions, privacy standards, and runtime execution restrictions.
• Trusted Models: Enables intelligent model routing across native and third-party models based on performance, cost, and task accuracy.

Centralized Visibility and Open Integration

To help management oversee expanding AI deployments, Salesforce is introducing a new AI Control Plane. The management layer provides centralized visibility into agent performance, access permissions, identity verification, behavior, and operational costs across both Salesforce and external third-party AI systems.

The framework is built with a headless architecture supporting standard protocols—including Model Context Protocol (MCP), APIs, Skills, and Plug-ins—allowing enterprise capabilities to integrate directly into external environments like Claude, Slack, and Microsoft Teams.

“AI is moving faster than any technology we’ve seen — which is exactly why composability matters so much to us. We don’t want to bet our future on one closed stack; we want the freedom to adapt as the landscape shifts. That’s what resonated with us about Salesforce’s new Enterprise AI Harness,” said Shawn Malhotra, CTO at Rocket Mortgage.

Availability

Many foundational technologies powering the Enterprise AI Harness are available immediately. The unified experience and new capabilities are scheduled to roll out beginning in early fiscal FY28, with existing Salesforce clients offered upgrade paths for eligible investments. Specific pricing, packaging, and regional availability details will be released closer to general availability.