Global CRM leader Salesforce has unveiled a new Trusted AI foundation designed to help businesses scale accurate, explainable, and secure AI across all workflows.

The announcement, made on Monday, introduces a unified framework called the “Agentic Enterprise,” where humans and AI collaborate seamlessly in decision-making and operations.

Salesforce said the new system will ensure AI outputs are grounded in unified business data, embedding transparency, governance, and compliance into every process.

Among the key innovations are Data Cloud Context Indexing, which allows AI agents to interpret complex documents; Data Cloud Clean Rooms, enabling secure data collaboration without exposing raw data; and Tableau Semantics, which translates raw data into business language for consistent insights.

The company is partnering with Databricks, dbt Labs, and Snowflake to standardize semantics across platforms. It also plans to acquire Informatica by early FY27 to strengthen metadata management and data governance.

Salesforce’s MuleSoft Agent Fabric will act as a central hub to register and orchestrate AI agents, supported by integrations with CrowdStrike and Okta for security and risk management.

Rahul Auradkar, Salesforce’s EVP and GM of Unified Data Services, said the foundation aims to harmonize structured and unstructured data to unlock intelligent automation and actionable insights.

“AI must be grounded in a deep business context to deliver true transformation,” Auradkar said. “With this foundation, we’re enabling our customers to scale responsibly and unlock the full potential of the Agentic Enterprise.”