Pipefy’s latest survey found a widening gap between AI adoption and AI orchestration among U.S. businesses: 67.1% of respondents said they use AI in at least some processes, but just 6.6% have fully orchestrated workflows from end to end.
As companies expand their use of AI agents, that gap could create new opportunities for MSPs and integrators to help customers connect systems, establish governance controls, and determine where human oversight is still required.
AI adoption is outpacing process orchestration
Pipefy surveyed 148 companies from its customer base in the United States and Brazil during the first half of 2026 using a structured questionnaire. The findings highlighted in the announcement reflect the U.S. portion of the survey.
The results point to continued fragmentation in how AI is incorporated into business processes. Pipefy reported that 42.6% of U.S. respondents still switch between multiple systems and screens to complete a single process.
“AI projects rarely stall because the model is wrong. They stall because the process around it was never built to be governed,” said Sobhan Daliry, chief product officer and AI strategy leader at Pipefy.
“An orchestration layer is what turns a pilot into an operation: it runs agents inside defined boundaries, with audit trails and human validation at the points where risk requires it.”
That orchestration gap could become more significant as organizations expand their use of AI agents. According to the survey, 41.3% of companies already use AI agents in internal processes, while another 41.3% plan to adopt them.
Human oversight remains central to AI workflows
Pipefy also found that human oversight remains part of respondents’ expectations for how AI-driven processes should operate.
The survey reported that 57.4% see humans serving as orchestrators and validators. Another 34.4% said they would delegate work to an AI agent if the rules governing its actions were clearly defined and auditable.
When evaluating AI purchases, ease of adoption was the most commonly cited consideration at 28%, followed by operational gains at 27% and clear business benefits and return on investment at 19%.
Pipefy said this suggests teams are prioritizing speed to value over the broader business case, with longer implementations more likely to lose internal support.
AI orchestration creates an opening for channel partners
Pipefy’s data points to a gap between using AI and actually embedding it into day-to-day operations. For partners, that could open more work around connecting systems and helping customers move beyond isolated AI use cases.
The rise of AI agents adds another layer. Customers may need help deciding where agents can act independently, where human approval is still required, and how those actions are tracked across multiple applications and processes.
That could give MSPs, integrators, and other service providers a larger role in the operational side of AI adoption. Instead of focusing only on deploying AI tools, partners may increasingly be brought in to make those tools work together in a governed, repeatable process.
Read More: Broadcom has agreed to provide Anthropic up to $42 billion in financing for AI infrastructure, deepening a relationship spanning chips, leasing, and compute.