Only 34% trust AI agents already deployed at 86%
AI agents have moved beyond the pilot stage at 86% of organisations surveyed, but only 34% of respondents said they trust the actions those systems are taking.
The figures come from a Forrester Consulting survey commissioned by Boomi, based on 409 director-and-above IT and technology decision-makers across North America, Europe and APAC.
Among organisations described as being in “agentic chaos” – the bottom quartile for operational readiness across governance, integration, API/MCP management and other categories – 77% are moving into production anyway. The study said those organisations face an average of $2.1 million in added costs from compliance fines, lost customers, operational downtime and rework.
Organisations with “agentic control”, defined as the top quartile for readiness, were much more confident. In that group, 55% reported high confidence in their agents’ actions and decisions, compared with 22% of organisations in agentic chaos.
Integration divides leaders from laggards
The biggest gap in the study was around integration. Organisations 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.
Use of integration platform as a service also split the two groups. It was used to support agentic workflows by 46% of organisations with agentic control, compared with 25% of those in agentic chaos.
When it came to building AI agents, 86% of organisations with agentic control said iPaaS and API management capabilities were important to readiness, compared with 58% of those in agentic chaos.
Steve Lucas, chairman and CEO at Boomi, said: “This research confirms what we’re seeing everywhere: the trust problem with agentic AI is really a data problem.”
The survey also found some organisations are running as many as 200 agents. Among leaders with agentic control, 46% had established central governance of MCP, compared with 32% in chaos, while 48% had aligned their AI and integration teams under one operating model, versus 37% of their peers in chaos.
Among organisations with agentic control, 59% reported productivity gains from their AI agents, 51% reported increased innovation, 46% said they gained reusable capabilities for other parts of the business, and 45% said they had automated repetitive tasks.
