5 Numbers From Camunda's 2026 Report Every Israeli CTO Should Know

In Camunda’s 2026 State of Agentic Orchestration & Automation report, 73% of respondents say there is a significant gap between their agentic AI vision and reality. The vendor’s survey covers 1,150 senior IT leaders, business decision-makers, and enterprise software architects. Camunda commissioned the research, conducted in September–October 2025 among large organizations in the US, UK, Germany, and France. These are self-reported findings from that sample, not a census of enterprises or an Israel-specific study. Here are five findings and our interpretation for an Israeli enterprise planning its next AI move.

1. 73% - the vision-reality gap

Nearly three out of four respondents report that gap. It is a useful signal that interest in agents and operational readiness are different things. The figure alone does not explain whether a particular project is held back by data, integration, model performance, budget, or governance.

What this means for your organization: if your project has stalled, identify the specific obstacle before buying more tooling. Compare the intended outcome with what the pilot has actually demonstrated.

2. 71% report agent use; 11% of use cases reached production

Camunda reports that 71% of organizations represented in the survey use AI agents, while 11% of agentic AI use cases reached production over the previous year. The denominators differ: organizations versus use cases. These figures do not tell us what share of organizations had at least one agent in production.

What this means for your organization: treat a successful pilot as evidence for the next step. Ask what remains to be proved with real data, real customers, production controls, and the expected workload.

3. 84% fear business risk without IT controls

In the survey, 84% cite business risk when AI operates without appropriate IT controls; 80% cite a lack of transparency about its use in business processes, and 66% cite compliance concerns. These concerns can overlap; the percentages should not be added together.

What this means for your organization: define who owns the outcome, what the agent is authorized to do, and when a person must intervene. Orchestration can enforce a defined sequence and approval gates, but those controls still need to be designed and checked against the requirements that apply to your organization.

4. 48% run AI agents in isolated silos

Camunda reports that 48% use agents in silos rather than within end-to-end processes. A locally useful agent may still leave handoffs, downstream work, and exception handling disconnected.

What this means for your organization: measure the whole process as well as the agent’s task. A faster step may have little effect on total turnaround time if the next queue remains unchanged. That is why we argue that agents need a conductor: someone must coordinate the steps before and after them.

5. 90% agree: AI needs orchestration like any other automated endpoint

Camunda reports that 90% of IT leaders surveyed agree that AI needs orchestration like other automated endpoints to support regulatory compliance, while 88% agree that orchestration across processes is needed to maximize the benefit of AI investment. These are respondents’ views, not measured proof of a return on investment.

What this means for your organization: there is strong support for orchestration within this sample. Turn that support into a bounded plan with an owner, baseline measures, and release criteria—the approach behind our 90-day orchestration roadmap.

What this means for your organization, overall

These findings suggest that governance and integration deserve attention alongside model capability. They do not establish a single cause of stalled adoption or prove that one platform solves it. For your organization, the useful next step is to test the operating model on one process before expanding agent use.

At NG Workshop, we help organizations in Israel turn that work into a process-first architecture with Camunda, clear human checkpoints, and evidence that operations and compliance teams can review.

Sources

• Camunda: survey methodology and launch findings
• Camunda: Closing the Agentic AI Vision-Reality Gap - 2026 State of Agentic Orchestration & Automation Report
• Camunda: Breaking Through the Automation Ceiling - CamundaCon 2026

Summary

• In the survey, 73% report a vision–reality gap; 71% report agent use, while 11% of agentic AI use cases reached production over the previous year.
• 84% cite risk without appropriate IT controls, and 48% report siloed agent use.
• 90% support orchestration for compliance; that opinion still needs to become a tested implementation plan.

Want to build a data-driven AI orchestration plan for your organization? Let’s talk