Salesforce has introduced Agentforce 3, an upgraded version of its digital labour platform aimed at providing enterprises with enhanced visibility and control for scaling AI agents across their operations.
Recent research and data cited by Salesforce highlights a significant increase in global AI agent adoption, reporting a 233% uptick in use over the past six months based on an upcoming Slack Workflow Index. During this period, more than 8,000 customers engaged with Agentforce.
Canadian market snapshot
New data from Salesforce points to widespread frustrations with current customer service experiences among Canadians. According to the research, 59% of Canadian respondents reported dissatisfaction due to unhelpful or non-existent self-service options, while 55% expressed negativity around excessive departmental transfers. Additionally, 45% felt they spent too much time obtaining the assistance they required.
The findings also indicate an openness among Canadians to engage with AI-driven customer service solutions. More than one in three surveyed stated a willingness to interact with AI agents if it would reduce wait times and avoid the need to consult traditional customer service representatives.
Agentforce 3 features
Agentforce 3 brings to market several technical updates. Among the key additions is the Command Center, designed to provide a comprehensive observability solution for managing AI agents. This tool aims to give leaders a single interface to monitor agent health, analyse performance, and optimise outcomes. Built into Agentforce Studio, the Command Center enables teams to derive insights across all AI agent interactions, uncovering usage patterns, identifying areas for improvement, and receiving AI-powered recommendations for various conversation types.
The system offers detailed live analytics, tracking latency, escalation frequency, and error rates, with automated alerts for anomalies. Dashboards enable users to monitor adoption, feedback, cost, success rates, and topic performance, with all agent activities recorded in a native, traceable data format within Data Cloud. Information from Agentforce integrates with monitoring tools including Datadog, Splunk, and Wayfound via the OpenTelemetry standard.
Agentforce 3 is equipped with support for Model Context Protocol (MCP), which facilitates plug-and-play interoperability, and introduces an expanded AgentExchange, providing access to prebuilt services from over 30 partners such as AWS, Box, Cisco, Google Cloud, IBM, Notion, PayPal, Stripe, Teradata, and Writer. The upgrade includes more than 100 new prebuilt industry actions and an updated Atlas architecture focusing on low latency, higher accuracy, resilience, and support for native large language models including Anthropic.
Management perspective and customer results
"With Agentforce, we've unified agents, data, apps, and metadata to create a digital labor platform, helping thousands of companies realize the promise of agentic AI today. Over the past several months, we've listened deeply to our customers and continued our rapid pace of technology innovation. The result is Agentforce 3, a major leap forward for our platform that brings greater intelligence, higher performance, and more trust and accountability to every Agentforce deployment. Agentforce 3 will redefine how humans and AI agents work together — driving breakthrough levels of productivity, efficiency, and business transformation," said Adam Evans, EVP & GM of Salesforce AI.
Salesforce reports that Agentforce has, in practical deployments, helped to decrease case handle times for clients such as Engine by 15%, and autonomously resolved 70% of 1-800Accountant's administrative chat exchanges during critical periods in 2025. Grupo Globo experienced a 22% improvement in subscriber retention through the platform, according to Salesforce figures.
The Command Center will initially support Service Cloud, surfacing agent activity in real-time dashboards for contact centre supervisors, with plans to extend departmental coverage. Development tools within Agentforce Studio allow natural language-driven agent building, topic generation, and testing, including AI-driven simulation for validating agent behaviour prior to deployment.
Agentforce 3's new capabilities are presented by Salesforce as addressing previously unmet needs for tooling, governance, and visibility required for enterprise-scale AI agent adoption.
