Salesforce introduced its Trusted Enterprise AI Harness on Friday, a common architecture the company says gives AI agents a shared understanding of the customer and the business so they can act inside enterprise controls.
Announced in Manila on September 11, the launch lands as agents take on more complex work across business functions. Salesforce frames the Enterprise AI Harness as its answer to a question enterprise buyers keep raising: how do you give agents what they need to work reliably, securely, and at scale without rebuilding those capabilities for every agent?
Six trusted capabilities
The harness groups six capabilities under one composable architecture: context, agency, action, governance, security, and models. Customers can run all six as one system or take only what they need, alongside Salesforce technology, their existing stack, or both — including third-party models, agents, and systems.
- Trusted Context brings customer data, metadata, semantics, knowledge, real-time signals, and memory together so agents understand the business around a customer.
- Trusted Agency covers reasoning, planning, state, and orchestration, pairing flexible AI reasoning with deterministic controls where certainty is required.
- Trusted Action connects agents to applications, APIs, workflows, and tools, moving them from understanding to doing.
- Trusted Governance applies lineage, quality, guardrails, and controls to the data and processes AI relies on.
- Trusted Security applies identity, permissions, privacy, and runtime security to what AI can access and do.
- Trusted Models routes work to the right model for each task based on accuracy, performance, cost, and business requirements.
What the Enterprise AI Harness looks like in practice
Salesforce uses one customer question to explain the architecture: “Can we fulfill this order today?” No single system holds the answer. CRM knows the customer relationship, ERP knows inventory, contracts hold entitlements, analytics supplies business definitions, and policies decide what can be promised.
The hard part is not connecting those systems but letting an agent understand them for one customer at one moment, then act securely within policy. Outcomes flow back into context, which Salesforce describes as a compounding intelligence loop.
“The Agentic Enterprise won’t be defined by which model a company chooses,” said Rohan Kumar, President, Chief Platform and Engineering Officer. “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.”
An AI Control Plane for the enterprise
Alongside the harness, Salesforce is introducing an AI Control Plane: one place to discover and register agents, set identity and policy, manage lifecycle, evaluate performance, observe behavior, and control cost across Salesforce and third-party AI.
The Enterprise AI Harness draws on Data 360, Informatica, MuleSoft, Agent Fabric, Tableau, Agentforce, Salesforce Guardian, and the Salesforce Platform. It is built headlessly, with capabilities reachable through MCP, APIs, Skills, and plug-ins, so they extend to where people already work — including 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,” said Shawn Malhotra, CTO, Rocket Mortgage.
Availability
Many of the technologies behind the harness are available today. New capabilities and the unified experience are planned to begin rolling out in early fiscal FY28, with upgrade paths for eligible existing Salesforce investments. Salesforce said it will announce details on availability, packaging, pricing, and upgrade paths closer to general availability, and noted that availability may vary by region and customer agreements.
Giancarlo Viterbo is a Filipino Technology Journalist, blogger and Editor of gadgetpilipinas.net, He is also a Geek, Dad and a Husband. He knows a lot about washing the dishes, doing some errands and following instructions from his boss on his day job. Follow him on twitter: @gianviterbo and @gadgetpilipinas.
Giancarlo Viterbo
Giancarlo Viterbo is a Filipino Technology Journalist, blogger and Editor of gadgetpilipinas.net, He is also a Geek, Dad and a Husband. He knows a lot about washing the dishes, doing some errands and following instructions from his boss on his day job. Follow him on twitter: @gianviterbo and @gadgetpilipinas.
View all articles →Related News
TCL Introduces New Smart Washer-Dryer Series
