Salesforce has unveiled Agentforce 360, the newest generation of its enterprise AI agent platform, introducing major new capabilities across voice, reasoning, and contextual grounding.

The release adds Agentforce Builder, a conversational development studio; Agent Script, a new scripting language for agent behavior; Agentforce Voice, a voice-first AI system for natural, on-brand conversations; and Intelligent Context, which grounds agents in complex, unstructured business data. The announcement signals Salesforce’s intent to dominate enterprise AI ahead of Dreamforce 2025, where automation and agentic systems are set to headline.

Salesforce’s Agentforce 360 is designed to help enterprises build AI systems that reason, adapt, and respond to complex business workflows. A highlight of the launch is Agent Script, a new tool that allows users to define “if/then” situations through text prompts — effectively giving AI agents rule-based flexibility in customer or operational scenarios. Agent Script will enter beta in November 2025, giving early users a chance to test adaptive prompting and conditional logic before full deployment next year.

The platform now supports “reasoning models” — AI systems capable of considering context before responding — developed in collaboration with Anthropic, OpenAI, and Google Gemini. Salesforce says these integrations enable agents to produce more reliable and contextually aware outputs, a key demand among enterprise clients.

Another major component of the upgrade, Agentforce Builder, serves as a unified environment to build, test, and deploy AI agents from a single interface. It includes Agentforce Vibes, a vibe-coding system unveiled earlier this month that lets developers tailor the “tone” and behavior of their enterprise apps.

The Agentforce Builder beta will launch in November 2025, bringing a simplified workflow for organizations scaling multiple AI projects across departments.

Agentforce Voice introduces a new voice layer for Salesforce’s AI ecosystem, designed to replace traditional IVR systems with fluid, real-time conversations that reflect each company’s tone and brand identity. Built with ultra-low latency and realistic speech synthesis, it enables AI agents to handle live interactions such as updating CRM records, triggering workflows, and escalating complex issues — all while maintaining auditability through Data 360.

Each voice agent can be tailored to a company’s brand, with customizable personality and data-trained context. Salesforce says it tested Agentforce Voice across thousands of real-world environments — from noisy backgrounds to multitasking callers — to ensure enterprise-grade reliability. The system integrates natively with Amazon Connect, Five9, Genesys, NiCE, and Vonage, and supports live transcription and “instant takeover” by human agents directly inside the Salesforce console.

Salesforce is deepening its AI footprint inside Slack, positioning the messaging platform as the company’s AI collaboration hub. Starting this month, Salesforce’s flagship apps — including Agentforce Sales, IT, and HR — will be directly accessible in Slack, with expanded functionality rolling out through 2026.

Slack is also piloting a redesigned Slackbot, evolving from a static chatbot into a personalized AI assistant that learns user habits and delivers actionable insights and suggestions. Salesforce plans to extend Slack’s reach with connectors to Gmail, Outlook, and Dropbox early next year, transforming it into a cross-platform enterprise search tool.

Salesforce is expanding Data 360 with Intelligent Context, a system that grounds AI agents in an enterprise’s unstructured data. The platform now activates a company’s structured and unstructured data to give every agent business context and personalization. The feature automatically extracts, structures, and surfaces information from complex sources like PDFs, diagrams, and other documents — transforming it into usable business knowledge that improves agent accuracy and relevance.

A low-code configuration pipeline allows organizations to process documents in hours rather than days, while multimodal support handles diverse content types such as tables, images, and flowcharts. This flexibility helps enterprises deliver context-rich, accurate interactions by ensuring every AI agent can reason with the right business information at the right time.

Data 360 now activates both structured and unstructured data to give every agent true business context and personalization. Through Tableau Semantics, Salesforce standardizes metrics and analytics across clouds using the Customer 360 Semantic Data Model (SDM) — developed in partnership with Databricks, dbt Labs, and Snowflake. This integration ensures consistent semantics and contextual understanding across Salesforce applications.

The enterprise AI race continues to accelerate. Just last week, Google announced Gemini Enterprise, a suite of tools for building large-scale AI agents with customers including Figma, Klarna, and Virgin Voyages. Meanwhile, Anthropic is gaining traction with Claude Enterprise, highlighted by its new deal with Deloitte and a strategic partnership with IBM to expand deployment globally.

Against this backdrop, Salesforce points to its 12,000 enterprise Agentforce customers — the largest user base of any AI agent platform to date — underscoring its momentum heading into Dreamforce 2025.

Salesforce is embedding Agentforce directly into its Customer 360 applications, turning each product suite into an AI-powered assistant. Agentforce Sales automates prospecting and quoting; Agentforce Marketing builds and launches campaigns; Agentforce Service delivers proactive, always-on customer support; and Agentforce Field Service integrates live mapping and hands-free data capture with Esri.

New additions include Agentforce Revenue Management, which enables agentic quoting and billing, and Agentforce IT Service, which replaces ticket-based systems with conversational resolution — accessible across Slack and 100+ partner integrations with Box, CrowdStrike, IBM, Okta, and Workday.

These features, combined with industry-specific versions such as Agentforce Life Sciences, Agentforce Public Sector, and Agentforce Manufacturing, position Salesforce as an AI platform that operates natively across departments and verticals.

Salesforce’s Agentforce 360 ecosystem is expanding rapidly through partnerships and integrations across cloud platforms. Customers can now deploy Anthropic’s Claude, Google’s Gemini, and OpenAI’s ChatGPT models directly within Salesforce via Amazon Bedrock and new Model Context Protocol (MCP) APIs.

The company also introduced Slack-native AgentExchange, a marketplace where organizations can discover and install AI agents from partners like Cursor, Perplexity, Vercel, and WRITER, all within Slack. This expands Slack’s role as the “Agentic OS” — a central hub where humans, agents, and data interact in real time across enterprise workflows.

Despite the momentum, Salesforce’s announcement arrives amid lingering skepticism about enterprise AI’s bottom line. A recent MIT study found that 95% of enterprise AI pilots fail before reaching production, underscoring the challenge of turning ambitious AI projects into measurable results.

Salesforce aims to create a 360-degree ecosystem for enterprise AI agents — a unified environment where reasoning, data, and collaboration converge to make AI adoption both scalable and practical for business operations.

Despite these challenges, Salesforce’s early adopters are already reporting measurable gains across industries.

Salesforce reports that more than 12,000 customers are already using Agentforce 360, with measurable results across industries. Reddit deflected 46% of support cases and reduced response times by 84%. Adecco automated half of its candidate conversations outside business hours, while OpenTable resolved 70% of diner and restaurant inquiries autonomously.

Other early adopters include Engine, which cut handle times by 15%, saving $2 million annually, and 1-800Accountant, which achieved a 90% case deflection rate during peak tax periods. Salesforce executives say these outcomes demonstrate how Agentforce 360 can drive operational efficiency and measurable ROI through agentic automation.

Agentforce 360 is available globally today, with continued innovation arriving in pilot and beta releases over the coming months. The upcoming Agentforce Builder and Agent Script betas launch in November 2025, with general availability extending into next year.

Salesforce encourages customers and developers to explore the full platform capabilities, integrations, and early-access programs at salesforce.com/agentforce360

Q1: What is Agentforce 360?
A: It’s Salesforce’s next-generation AI agent platform with new prompting, reasoning, and deployment tools for enterprise customers.

Q2: What are the standout new features?
A: Agent Script for flexible logic, Agentforce Builder for unified development, and expanded Slack integration with personalized AI assistants.

Q3: Who powers the reasoning models?
A: Anthropic, OpenAI, and Google Gemini provide the reasoning model architectures behind Agentforce 360.

Q4: When will the new tools be available?
A: Both Agent Script and Agentforce Builder enter beta in November 2025; full rollout extends into early 2026.

Q5: How does Salesforce compare to competitors?
A: With 12,000 customers, Salesforce’s Agentforce platform leads in enterprise adoption, outpacing recent launches from Google and Anthropic.

Q6: What new capabilities set Agentforce 360 apart?
A: Salesforce added Agentforce Voice, Intelligent Context, and expanded Data 360 integration, giving enterprises natural conversation, deep data grounding, and unified AI development tools.

The launch of Agentforce 360 marks more than a feature update — it’s Salesforce’s declaration that it intends to own the enterprise AI layer, not just integrate with it. By fusing reasoning models, custom agent scripting, and Slack-based collaboration, Salesforce is positioning itself as the operating system for AI-powered enterprises, where agents can act across departments and data sources in real time.

This strategy also signals a decisive shift toward vertical AI integration. Rather than relying solely on partners like Anthropic, OpenAI, or Google Gemini, Salesforce is building the connective infrastructure that ensures those reasoning models work seamlessly inside its own ecosystem. That control over the agent environment — from data to deployment — could redefine how businesses operationalize AI at scale.

At a market level, Agentforce 360 represents the next phase of enterprise AI maturity: from experiments to embedded systems. The combination of low-code agent creation, integrated reasoning, and conversational collaboration tools addresses the biggest friction points enterprises face — fragmentation, security, and ROI uncertainty.

If successful, Salesforce could establish a new category of enterprise agent platforms — systems that unify reasoning, workflow, and communication in one trusted environment. For organizations still struggling to make AI practical, Agentforce 360 offers a blueprint for how intelligence can finally move from pilot to production.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant used for research and drafting. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.