Boomi, the data activation company for AI, today announced new research conducted by Forrester Consulting on behalf of Boomi showing that despite rapid enterprise adoption of AI agents, trust hasn’t kept pace with ambition.
The Forrester survey of 409 director-and-above IT and technology decision-makers across North America, Europe, and APAC found that 86% of organizations have moved beyond the AI agent pilot stage, yet just 34% say they trust the actions their AI agents are taking. Among organizations in a state of “agentic chaos” (the bottom quartile for operational readiness across governance, integration, API/MCP management, and other categories), 77% are moving into production anyway, exposing themselves to an average of $2.1 million in added costs from compliance fines, lost customers, operational downtime, and rework. Organizations with “agentic control” (the top quartile for readiness) are far more measured, and far more confident: 55% report high confidence in their agents’ actions and decisions, compared with just 22% of those in agentic chaos.
“This research confirms what we’re seeing everywhere: the trust problem with agentic AI is really a data problem,” said Steve Lucas, Chairman and CEO at Boomi. “Agents can only be trusted to act on data that’s been properly activated, connected, and governed, and most companies deployed agents before they did that work. The ones who did it first are the ones getting real value now.”
Integration Is What Builds Trust
The research identifies integration as the clearest line between enterprises that trust their agentic AI and those that don’t. Decision-makers with agentic control were three times as likely as those in agentic chaos to say reliable, well-managed APIs determine whether they pilot a use case at all. Integration platform as a service (iPaaS) showed the widest adoption gap of any method surveyed: 46% of organizations with agentic control use iPaaS to support agentic workflows, compared with just 25% of those in agentic chaos.
The gap is even starker when it comes to building the agents themselves. Eighty-six percent of organizations with agentic control say iPaaS and API management capabilities for building AI agents are important to their readiness, compared with just 58% of those in agentic chaos: the single widest gap Forrester measured in the entire study. As Forrester puts it in the study, “If LLMs are like brains, iPaaS products are a limb that empowers that brain to act upon the physical world.”
Leaders with agentic control were also far more likely to prioritize the operational work behind the scenes: 47% cite improving integration with tools, APIs, and apps as a top focus area, versus 31% of those in chaos, who remain focused on improving the AI model itself without building the connections needed to act on its decisions.
That same gap shows up in how organizations are handling agent sprawl. Some are now running as many as 200 agents, a symptom of what Forrester calls “POC/pilot purgatory,” where high ambition for AI agents stalls out because those agents were never connected to the enterprise systems they’d need to act on their decisions. Leaders with agentic control are far more likely to get ahead of this: 46% have established central governance of MCP, the standard that lets agents connect to enterprise systems safely, compared with 32% in chaos, and 48% have aligned their AI and integration teams under one operating model, compared with just 37% of their peers still in chaos.
The Payoff
The organizations investing in a strong integration layer are the ones seeing it pay off. Among those with agentic control, 59% reported productivity gains from their agentic AI deployments, 51% reported increased innovation, 46% gained reusable capabilities they could apply elsewhere in the business, and 45% automated repetitive tasks.
Forrester recommends that organizations looking to close the trust gap align their AI and integration teams under one operating model, implement a control plane to govern AI agents, and introduce an orchestration layer that gives agents the reach to connect to data, applications, and each other.
