EMEA Executives Struggle with AI Vendor Lock-in and Visibility, IBM Reports

A new study from IBM indicates that digital sovereignty is a growing priority for EMEA businesses, yet many are locked into AI systems they struggle to control. [1]

According to the report, 90% of surveyed EMEA executives do not fully understand their organization's dependencies across AI models, vendors, and infrastructure. [1] Furthermore, nearly three-quarters of respondents reported that switching their primary AI model or vendor would be difficult. [1] This combination of limited visibility and vendor lock-in poses significant operational risks, with 81% of executives stating that a seven-day outage at their primary AI provider would have a severe or critical impact on their organization. [1]

Speaking at IBM's AI Summit in London, Gregory Verlinden, vice president of data and AI at Cegeka, noted that the costs associated with vendor lock-in can be invisible until it is too late. [1] Verlinden stated that AI token consumption is emerging as one of the largest expense categories after labor. [1]

He suggested that prioritizing a flexible, multi-vendor environment can provide a more resilient framework and help reduce dependence on any single provider. [1] Additionally, Nina Wilhelmsen, an IBM EMEA business lead, argued that governing an increasing number of AI agents requires greater organizational visibility. [1]

The high percentage of executives lacking visibility into their AI supply chain highlights a critical vulnerability as enterprise AI adoption scales. The fact that nearly three-quarters of firms cannot easily switch vendors indicates that primary AI providers currently hold significant leverage over their customers. Compared to traditional IT supply chain disruptions, the rapid emergence of AI token costs as a major expense could quickly transform these dependencies into severe financial liabilities for enterprises that remain locked into single-provider ecosystems. What the sources don't address: Whether this lack of visibility is primarily due to the rapid pace of AI adoption or a fundamental lack of governance tools available in the current market.