Salesforce backs open AI agent ecosystem with interoperability push

Salesforce has unveiled a major update to its digital labour platform, Agentforce, introducing native support for Model Context Protocol (MCP) and announcing pilot support for MCP servers on AgentExchange, its AI agent marketplace.

The update marks a significant move toward trusted agent interoperability, allowing partners to build and connect AI agents across platforms without sacrificing governance or enterprise security.

Solving the AI paradox

While AI holds the promise of transforming operations with intelligent digital agents, businesses often face a dilemma—choose between innovation and secure control. Closed ecosystems offer safety but restrict flexibility. Open systems foster innovation but raise compliance concerns.

Salesforce’s new architecture seeks to eliminate this trade-off by offering a trusted open ecosystem for interoperable AI agents, accelerating enterprise adoption while maintaining tight governance.

What's changing with Agentforce

The introduction of MCP and upcoming Agent2Agent (A2A) support simplifies AI agent development for partners. With this model:

• AI agents can now plug into enterprise systems out-of-the-box

• Integration no longer demands one-off APIs or custom builds

• Partners can deploy, monetise and govern their agent solutions at scale

• Enterprise security standards are enforced across the ecosystem

Salesforce claims this new setup allows organisations to deliver agentic solutions up to 16 times faster than traditional DIY approaches.

Over 30 partners, including AWS, Google Cloud, IBM, Box, PayPal, Stripe, and WRITER, have committed to listing MCP servers on AgentExchange. These integrations allow Agentforce to pull in real-time data, process documents, handle commerce flows, manage contracts, and power content generation—without building from scratch.

Examples of use cases include:

• AWS: Agents processing unstructured data, querying knowledge bases, and generating insights from video, audio, and code repositories

• Google Cloud: Connecting with generative AI models like Imagen and Lyria, Maps, and transactional databases for seamless workflows

• Stripe & PayPal: Handling payments, disputes, subscriptions, and refunds through AI-driven automations

• WRITER: Enabling AI-generated content and compliance checks directly inside Salesforce

In a strategic move, Salesforce has merged AppExchange and Global Partnerships under a single CEO, Brian Landsman, to streamline innovation and partner collaboration.

“The focus is on enabling partners to reuse their IP, build once, and connect everywhere,” Landsman said. “We want customers to move away from one-off agent builds and toward a composable, trusted ecosystem.”

Salesforce is also working with system integrators and consulting partners to guide enterprises through their agent transformation—from proof of concept to full-scale rollouts. A growing pool of 272,000 AI-certified specialists is already supporting this shift.

Salesforce Ventures and Global Partnerships are also developing a dedicated program to support next-gen AI startups. The initiative will provide:

• Early access to Agentforce infrastructure

• Technical mentorship

• Streamlined routes to integrate with AgentExchange

• Access to Salesforce’s global customer network

This is expected to unlock faster time-to-market for emerging companies working on enterprise-grade AI tools.

As enterprises race to integrate AI into their operations, Salesforce’s open ecosystem approach offers a scalable and secure way forward. By blending interoperability with governance, the company is positioning Agentforce as the foundation for the $6 trillion digital labour market.

The focus now shifts from building isolated AI tools to deploying cohesive agent systems that are ready to interoperate, scale, and deliver measurable outcomes.

Read more : 

AMD and HCLTech team up to build enterprise tech for AI and cloud era

Sinch launches unified communication platform for Indian enterprises

Akamai closes API security gaps with lifecycle visibility enhancements

Tariffs and AI Drive burnout in India’s C-Suite