Salesforce has announced Agentforce 360, making its agentic AI platform generally available worldwide as competition around autonomous agents in the enterprise software space intensifies.
The company stated that the platform connects people, AI agents, and data within a single, trusted system and added that it debuts at Dreamforce 2025.
Agentforce 360 serves as a unifying layer across Salesforce’s Customer 360, allowing organizations to build, deploy, and govern AI agents for sales, service, marketing, and IT workflows.
Furthermore, Salesforce confirmed that the platform is live globally, and it emphasized features designed to ensure enterprise control, reliability, and security. According to the company, Agentforce 360 includes a conversational “Agentforce Builder” for designing and testing agents, a native “Agentforce Voice” layer for natural-language telephony, and hybrid reasoning with guardrails to balance deterministic workflows with large-language-model flexibility.
In addition, the platform incorporates observability dashboards to monitor reasoning quality and compliance.
Salesforce also noted that Slack will integrate conversational AI, enabling employees to pull data and trigger tasks within chats while maintaining enterprise-grade controls. Moreover, Salesforce highlighted an ecosystem approach that brings partner-built agents and actions into Slack to extend its capabilities.
“Agentforce 360 has reportedly been adopted by more than 12,000 customers,” the company said, citing brands such as Reddit, OpenTable, and The Adecco Group.
The announcement comes as large enterprise vendors accelerate their agent strategies. For example, Microsoft has added multi-agent orchestration to Copilot Studio, ServiceNow has promoted an “any AI, any agent, any model” platform, and Google and Oracle have expanded access to Gemini models for agentic workloads. These developments underscore how rapidly agent frameworks have moved from pilot programs to broad enterprise deployment.
