Salesforce unveils Agentforce 3 for enterprise AI agent control and scale

Salesforce has officially rolled out Agentforce 3, the latest iteration of its digital labour platform, delivering a significant boost to how enterprises deploy, manage, and scale AI agents. The upgrade addresses some of the most pressing challenges faced by organisations embracing AI at scale—lack of visibility, slow agent iteration, and fragile integration into existing enterprise systems.

Agentforce 3 is built on insights gathered from thousands of real-world deployments since its original launch in October 2024. The result is a platform that not only powers high-performance AI agents but also equips businesses with the observability and governance necessary for enterprise-wide transformation.

The Command Centre: A New Standard in Agent Observability

At the heart of Agentforce 3 is the new Agentforce Command Centre, designed to give organisations a real-time, unified view of AI agent performance. Teams can track critical metrics such as latency, escalation frequency, and error rates, while also drilling down into specific interactions to identify patterns and areas for optimisation.

With native integration to monitoring tools like Datadog, Splunk, and Wayfound, companies gain end-to-end visibility across both AI and human workflows. Salesforce’s use of OpenTelemetry standards ensures compatibility with existing observability stacks, streamlining adoption.

Agentforce 3 introduces native support for Model Context Protocol (MCP), allowing AI agents to interact with enterprise systems securely and without custom code. Dubbed the "USB-C for AI," MCP enables seamless connectivity between agents and business tools, governed by existing identity and security policies.

Salesforce is also leveraging its MuleSoft and Heroku platforms to convert any API into an MCP server or host custom MCP actions with minimal developer effort. This means organisations can rapidly orchestrate multi-agent workflows that connect to a variety of internal and external systems.

With the launch of Agentforce 3, Salesforce is significantly expanding its AgentExchange ecosystem, integrating over 30 new MCP-enabled partners. These include AWS, Google Cloud, Stripe, PayPal, and WRITER, offering AI agents access to a wide range of business actions such as document processing, payment handling, database queries, and content generation.

This expansion ensures that AI agents are not only smarter but also more useful—equipped with capabilities that accelerate business outcomes from day one.

Agentforce 3 builds on Salesforce’s Atlas architecture, delivering faster performance, reduced latency, and improved resilience. New enhancements include automatic model failover, response streaming, and support for additional languages such as Japanese, Portuguese, and German.

To meet the demands of regulated industries, Salesforce is also hosting Anthropic's Claude model within its trust boundary, with plans to add support for Google Gemini later this year. The platform has been deployed globally, now reaching Canada, the UK, India, Japan, and Brazil, and is authorised for FedRAMP High in the US public sector.

With Agentforce 3, Salesforce is redefining how organisations adopt and scale AI agents. By blending real-time observability, secure interoperability, and a rich ecosystem of prebuilt actions, the platform equips businesses with everything they need to unlock AI-driven productivity at scale.

As AI agents shift from experimentation to enterprise norm, Agentforce 3 positions Salesforce as a clear leader in trusted, agent-first automation.

read more :

Decoding the system integrator transformation in India

International women in engineering day 2025 - Shaping up the future

eScan XDR Unveils next-gen phishing simulator to tackle AI-powered scams

Adobe Creative Cloud announces AI-Driven updates across photoshop