As AI moves from experimentation to enterprise-wide deployment, two-thirds of CIOs and CTOs surveyed report being responsible for AI systems they don’t fully control, while governance is struggling to keep up at scale, according to new research from the IBM Institute for Business Value.
Also read: AiThority interview with Matej Bukovinski, Chief Technology Officer, Nutrient
A global survey* of 2,000 C-level technology executives (technology CxOs) found that the lack of visibility is widespread. The majority of executives surveyed (70%) say teams across the enterprise are adopting technology faster than IT departments can track it.
At the same time, technology leaders face increasing pressure to scale AI faster, even though many companies lack the structures to support it. Chief technology executives surveyed expect the number of AI agents deployed to increase by 38% by 2027. Although 80% of respondents report a mandate for CEO-led AI transformation, only 11% believe they are fully prepared for the scale of AI agent adoption expected next year. Governance is also lagging behind, with 77% of organizations surveyed reporting that AI adoption has already outpaced their current governance capabilities.
“For CIOs and CTOs, the challenge now is to scale AI systems that operate continuously and autonomously, often within governance models and architectures designed for slower, more predictable environments,” he said. Matt Wrightson, CIO, IBM. “It’s no longer just about accelerating AI adoption; we’re redesigning how organizations control, manage, and invest in AI, building in control and visibility from the start so they can scale with confidence.”
As AI scales, operational and security risks increase
- Our analysis shows that organizations that rely on manual governance experience an increased risk of incidents as their AI deployment expands, while organizations that embed controls directly into their AI systems experience a 25% reduction in incidents.
- Most technology CxOs surveyed (59%) cited security and compliance concerns as the biggest barrier to scaling AI agents.
- Organizations surveyed experienced an average of 54 AI agent incidents in the last year. In this incident, an unintended or harmful event occurred that required human remediation.
- According to respondents, 17% of reported AI agent incidents were high severity and took more than four hours to contain.
- 37% led to a data breach or security breach
- 33% caused cascading system failures
- 17% caused compliance issues
Organizations that redesign control and invest in AI can achieve stronger outcomes
- Spending on AI is expected to increase from just under 15% of IT budgets in 2025 to nearly 25% by 2027, a 71% increase in two years, increasing the risks for CIOs and CTOs.
- However, 84% of technology company CEOs do not fully operationalize AI financial management, and 85% still lack full visibility into real-time AI spending.
- Our analysis found that organizations that embed controls in their AI systems:
- Deploy 16x more AI agents than those relying on manual governance
- Achieved 18% high operating profit margin
- Cut your AI budget by a quarter
- Analysis shows that the organization has strong financial discipline.
- Deploy 2.4x more AI agents without increasing your AI/IT budget
- 3x more likely to say they are completely ready for AI at scale
- Organizations surveyed reported a 10% increase in AI return on investment by 2025 by designing for adaptability early, keeping workloads portable and models replaceable rather than locking in hard dependencies.
