Major enterprises continue to channel substantial capital into artificial intelligence despite broader economic turbulence, yet measurable business value remains elusive for most. Dun & Bradstreet surveyed 10,000 businesses across 32 countries for its Global AI Momentum Survey. The findings reveal that 65 percent of respondents intend to expand their AI budgets over the next twelve months. However, commercial returns have stalled, with only 28 percent of organizations reporting a noticeable or strong financial return on their technology investments to date.
The primary impediment to scaling initiatives lies within existing corporate data infrastructure, which frequently fails to meet enterprise requirements. A mere 8 percent of enterprises surveyed by Dun & Bradstreet describe their data architecture and quality as fully ready for organization-wide production environments. Another 33 percent consider their data mostly prepared, whereas 44 percent acknowledge significant structural deficits. Without clean, integrated data assets, even advanced foundation models fail to deliver consistent results or sustain automated business operations reliably.
At the same time, operational pressure on leadership teams is escalating due to failures in everyday production workflows. A parallel study conducted by IBM among more than 3,200 executives across eight EMEA countries reveals that 80 percent of companies saw at least one critical AI incident reach the board or C-suite over the past year. Reported issues included erratic agent behavior, erroneous automated decisions, and accidental data disclosures. These developments indicate that machine intelligence has exited the experimental realm and now poses tangible liability risks at the highest executive levels.
Despite rising awareness of operational hazards, institutional remedies remain surprisingly sluggish across European, Middle Eastern, and African corporations. According to IBM, only 40 percent of affected enterprises revised their internal governance and compliance policies following an escalated failure. Moreover, 56 percent of organizations continue to accelerate deployment even when existing governance structures are known to be unprepared. Only 37 percent maintain a quantified risk assessment to measure the financial liabilities linked directly to governance omissions.
Taken together, the empirical findings illustrate an enterprise landscape struggling to manage the transition from superficial pilot phases to industrial operational scale. The market pressure to demonstrate rapid progress frequently eclipses foundational tasks such as pipeline engineering, access control, and incident response planning. Companies that purchase software licenses without establishing solid data hygiene and governance safeguards risk misallocating capital while exposing themselves to avoidable liabilities. Sustainable competitive advantage in this environment will belong to organizations that prioritize disciplined data management and verifiable oversight over sheer deployment speed.