72 Percent Of Organizations Say Process Problems Caused AI Failures, Costing $1.55 Million On Average

A new report from Camunda suggests many organizations are attempting to incorporate AI tools into workflows that were designed long before AI became part of day-to-day operations.

New research from Camunda suggests organizations may be dedicating too much budget on artificial intelligence and not enough on redesigning the business processes that support it.

Enterprises have spent the last two years racing to deploy generative and agentic AI technologies. But according to a new survey commissioned by Camunda, the biggest obstacle to AI success may not be the technology itself.

The company's report found that 72 percent of organizations say process-related challenges have caused AI initiatives to fail, with respondents estimating those failures cost an average of $1.55 million per organization.

The findings suggest a growing disconnect between AI investment and organizational readiness. While businesses continue to invest heavily in AI, infrastructure and software, many organizations are attempting to incorporate these tools into workflows that were designed long before AI became part of day-to-day operations.

"Organizations are rapidly adopting AI. But they are applying it to processes designed for a world before AI, then wondering why the return on investment falls short," said Kurt Petersen, senior vice president of customer success at Camunda, in a news statement.

Organizations See A Need To Rebuild Processes

According to the report, 78 percent of respondents said their organization needs to fundamentally redesign business processes to remain competitive rather than simply later AI on top of existing workflows. Seventy-nine percent said that adding AI to existing processes creates less organizational resistance than undertaking a full redesign.

The survey also found that 82 percent of organizations believe their AI investments will ultimately fail without greater investment in process redesign. More than 80 percent said adapting their most important business processes for AI could take as long as five years.

For midmarket IT leaders, the findings highlight a recurring challenge surrounding AI initiatives. Bolting AI onto outdated workflows can create inefficiency rather than create it by automating processes that were already broken or over-complicated.

Leaders And Employees Don't View AI The Same

The research also revealed a disconnect between leadership's perceptions about AI and employee experiences.

While 89 percent of organizational leaders said AI is making teams more productive, only 64 percent of employees agreed. Moreover, 44 percent of employees said they have manually overridden AI-generated outputs because the underlying business process was not set up correctly.

Other findings suggest that employees often feel excluded from AI implementation decisions. Almost 70 percent reported they were not fully consulted on how AI would be introduced into their daily work. Sixty-five percent said they could use AI more effectively if implementations were handled differently.

Such employee sentiment appears to have business impact. Nearly half of respondents said their organizations have rolled back AI initiatives because of negative effects on employees' ability to perform their jobs.

AI Governance Problems Continue To Grow

The report also revealed governance and compliance concerns tied to AI adoption.

According to the survey, 40 percent of organizations experienced an AI-related compliance or governance issue during the past year. Among them, 84 percent said process challenges contributed to the issue.

Ninety-six percent of employees said they expect business process challenges to contribute to future AI-related compliance issues. And nearly half of organizations expressed concerns about AI systems acting outside expected boundaries if processes are changed.

Based on the report, midmarket IT leaders may need to weigh how much focus (and budget) to allocate redesigning the processes AI will ultimately support. The path they take may ultimately determine if AI investments truly deliver business value or become another expensive technology initiative with disappointing returns.