Multi-Agent Adoption to Surge 67% by 2027 as Enterprises Race Toward Agentic Transformation — Unified Architecture Key to Success

As agent adoption hits critical mass, 96% of IT leaders say AI agent success depends on integration across systems

IT leaders are turning to API-driven architectures to connect multi-agents and prevent fragmented infrastructure and shadow AI from stalling innovation

The transition to an Agentic Enterprise, where humans and AI agents work together, is gaining momentum as organizations currently use an average of 12 agents, with the number projected to climb 67% within two years. Already, 83% of organizations report that most or all teams and functions have adopted AI agents.* However, IT leaders face looming orchestration and governance challenges: 50% of agents currently operate in isolated silos versus part of a multi-agent system, resulting in disconnected workflows, redundant automations, and the potential risk of shadow AI.

To address these issues, the research - based on a survey of 1,050 enterprise IT leaders - also uncovered that respondents are turning to API-driven architectures as a unified foundation to connect, orchestrate, and govern multi-agents and drive AI success.

The Road to Multi-Agents

As adoption hits critical mass, AI agents are no longer experimental - they are becoming the primary driver of enterprise productivity. IT leaders are focused on using diverse agentic solutions and establishing agent communication protocols to manage their fleet of agents.

• High expectations: 96% of IT leaders say agents already have improved or that they expect them to improve employee experiences, and 95% believe they will free developers to focus on higher-value work.
• Diverse development: On average, organizations report that their existing AI agents were developed through various methods, split across:
  • Prebuilt SaaS agents (36%)
  • Embedded agents within enterprise platforms (34%)
  • Custom-built in-house (30%)
• Protocol adoption: As organizations deploy AI agents, they are actively supporting, or planning to support, a range of standards or protocols to manage and connect them, with high levels of interest in:
  • Agent Network Protocol (43%)
  • Agent Communication Protocol (43%)
  • Agent-to-Agent Protocol (40%)
  • Model Context Protocol (39%)
  • Universal Tool Calling Protocol (34%)

The Orchestration and Governance Gap

A critical orchestration and governance gap is emerging as enterprises race to deploy AI agents everywhere. While adoption is high, the infrastructure supporting it needs to be more integrated to support a multi-agent workforce that can collaborate and securely leverage data from across the enterprise.

• App and agent sprawl: The number of apps in enterprises grew from 897 to 957 year over year, with only 27% of them integrated together. With integration challenges and agent silos, 86% of IT leaders are concerned that agents will introduce more complexity than value.
• Top hurdles: The primary challenges currently hamper agentic transformation:
  • Risk management, compliance/security, and/or legal implications (42%)
  • Lack of internal expertise in AI/agent design (41%)
  • Legacy infrastructure or system incompatibility (37%)
  • Integrating siloed apps and data (35%)
• Data barriers: 96% of organizations experience barriers to using data for AI use cases, with 40% identifying outdated IT architecture/infrastructure due to data silos /disconnected systems as a top blocker.
• Rise of shadow AI: Nearly half (49%) of organizations cite cross-application data governance as a top integration challenge. An estimated 27% of APIs are currently ungoverned, on average, and only 54% of organizations have a centralized governance framework with formal oversight for their agentic capabilities.

Building a Unified Foundation

To bridge the integration gaps, IT leaders are moving toward a unified foundation. By using APIs as the "connective tissue," organizations can transform fragmented AI tools into a cohesive, multi-agent system where agents can safely communicate, share data context, and execute tasks across the entire IT estate.

• Connectivity mandate: 96% of IT leaders agree that AI agent success depends on seamless data integration across all systems.
• Architecture shift: 94% of IT leaders agree that AI agent success will require IT architecture to become more API-driven, where APIs are fundamental building blocks for connecting applications, data, and AI across an enterprise.
• Accelerating integration: One-third (33%) of teams are already leveraging APIs to speed up integration across systems.
• APIs for AI: 50% of organizations are already using APIs to connect and govern AI today.

Multi-Agents in Action

Early AI agent adopters are already demonstrating how a unified foundation can move agents from experimental silos into core business operations where multiple AI systems can work together:

• AstraZeneca, a global, science-led biopharmaceutical company, selected Agentforce Life Sciences for Customer Engagement to help transform its customer engagement globally, fostering stronger relationships with healthcare professionals (HCPs) through data-driven, AI-powered engagement. Extending its composable architecture with MuleSoft Agent Fabric, AstraZeneca will orchestrate internal and external agent actions across field engagement, commercial operations, and different brands and regions, allowing its care teams and AI agents to work seamlessly together.
• r.Potential, an enterprise intelligence company that delivers strategic insights to help executives optimize their workforce, is using Salesforce's core platform and Agentforce to develop its solutions. And with MuleSoft Agent Fabric, the company has the secure and governed foundation to orchestrate Agentforce, other specialized agents, and MCP tools to analyze external context and proprietary data and generate C-level insights.

The true success of an Agentic Enterprise isn't found in the sheer number of agents deployed but the overall effectiveness of those agents.

Perspectives:

• "The true success of an Agentic Enterprise isn't found in the sheer number of agents deployed but the overall effectiveness of those agents. We need to think about how they are discovered, governed, and orchestrated to work together. As we move into this multi-agent era, the role of IT is evolving from managing silos to building a unified foundation as the central control plane that allows multi-agent systems to be safe, reliable, and scalable." - Andrew Comstock, SVP and GM, MuleSoft, Salesforce
• "This year's Salesforce and Deloitte Digital research findings highlight a critical inflection point where organizations must move from simply deploying agents to operationalizing them at scale. Success requires reimagining integration strategies to build a foundation that is sustainable and secure. By establishing API-driven guardrails, enterprises can ensure their agentic transformation is ready for the demands of the modern enterprise." - Kurt Anderson, Managing Director and API Transformation Leader, Deloitte Consulting LLP

More information:

Methodology: Salesforce's 11th annual Connectivity Benchmark Report, in collaboration with Vanson Bourne and with insights from Deloitte Digital, uses survey data from interviews with 1,050 IT leaders across the globe. We conducted this double-anonymous online survey between October and November 2025 across the United States, the United Kingdom, France, Germany, the Netherlands, Australia, Singapore, Hong Kong, and Japan. We ensured that only suitable participants responded to the survey using a rigorous, multilevel screening process. Respondents are all IT leaders with managerial positions or above in an IT department. All respondents work at an enterprise organization (defined as having at least 1,000 employees) in the public or private sector.

Please see www.deloitte.com/us/about for a detailed description of their legal structure.

*AI agents are defined as the newest type of AI that can understand natural language, reason about goals, make decisions, and carry out tasks on its own. Agentic AI can coordinate with people, systems, or even other agents to achieve outcomes without step-by-step human input. Examples: virtual assistants that handle complex support cases end-to-end, AI tools that schedule meetings and send follow-ups, or systems that monitor operations and take corrective actions automatically.