70% Admit ‘Incomplete’ Understanding of AI Data Requirements While 21% Have ‘Insufficient’ or ‘Zero’ AI Control
Couchbase, Inc. (NASDAQ: BASE), a leading developer data platform tailored for AI-driven applications, has released the findings of its eighth annual survey of global IT leaders. The research, conducted in April 2025, involved 800 senior IT decision-makers from organizations with over 1,000 employees, spanning industries such as finance, healthcare, gaming, and technology. The results underscore the pressing need for organizations to move quickly and strategically in their AI journeys—or risk substantial financial and operational losses.
According to the survey, enterprises that fail to deploy AI effectively and in a timely manner risk losing an average of 8.6% of their monthly revenue. This translates to nearly $87 million in annual losses per company, based on the average revenue data from respondents. The consequences are particularly alarming given the scale of opportunity: 78% of IT leaders believe that early adopters of AI will set the pace for their industries, while 73% confirm that AI has already started transforming their organization’s technology environment.
However, widespread “decision paralysis” is stalling progress. An overwhelming 64% of respondents admit that internal hesitations and uncertainty are preventing their organizations from fully capitalizing on AI. Meanwhile, 21% of enterprises say they have either no control or inadequate oversight of AI usage, with employee access either too restricted or too open—an imbalance that introduces compliance risks, security vulnerabilities, and wasted resources.
AI Spending to Soar Amid Growing Urgency
The survey reveals a notable shift in digital priorities. AI-related spending, including investments in Generative AI (GenAI), agentic AI, and other forms of intelligent automation, is projected to surge by 51% between 2025 and 2026. This represents a sharper increase compared to the 35% growth forecast for overall digital modernization efforts. By 2026, AI investments are expected to make up more than half of all digital transformation budgets—evidence of the technology’s growing influence on business strategy.
Julie Irish, Chief Information Officer at Couchbase, emphasized the critical role of data in achieving AI success. “The evolution from GenAI to agentic AI is unlocking vast new opportunities. But without the right data foundation—one that prioritizes quality, accessibility, and scalability—enterprises will struggle to develop and scale impactful AI applications,” Irish said. “It’s not enough to experiment. Leaders must ensure they are operationalizing AI with the right controls and architecture in place.”
Data Infrastructure Struggles Hamper AI Projects
The report highlights deep-rooted infrastructure challenges that limit enterprise readiness for AI deployment. Notably:
- 99% of enterprises have faced disruptions in AI projects—whether due to poor data access, inadequate management tools, or cost overruns. These setbacks consume roughly 17% of their AI budgets and delay strategic initiatives by six months on average.
- 70% of organizations admit to having an incomplete understanding of the data required for AI to function optimally. This data gap contributes to the 62% that lack visibility into where their organizations may be vulnerable to AI-related risks, such as security breaches or compliance failures.
- Data architecture itself is often a limiting factor. The current average lifespan of an enterprise’s AI-supporting architecture is just 18 months, after which it must be overhauled or replaced.
- 75% of companies rely on multi-database environments, which complicate efforts to ensure data consistency across AI systems.
- 84% of enterprises lack the tools to store and manage high-dimensional vector data, which is essential for powering many modern AI applications.
- Additionally, 61% cannot guarantee that proprietary data used in AI training will not be shared externally, raising significant regulatory and data sovereignty concerns.
In response, every enterprise surveyed reported active efforts to simplify and consolidate their AI tech stacks, aiming for streamlined control, better scalability, and more secure operations.
Organizational Culture Plays a Key Role
The survey found that corporate mindset significantly affects AI outcomes. Enterprises that promote a culture of experimentation and innovation experience 10% more AI projects reaching production and 13% less AI investment waste than those with more restrictive or risk-averse approaches. Flexibility, cross-departmental collaboration, and a willingness to test new use cases all contribute to greater returns on AI initiatives.
Interestingly, spending on agentic AI is now on par with GenAI and other AI forms. Agentic AI—systems capable of goal-directed behavior without continuous human input—accounts for 30% of total AI spending, while GenAI and other forms each comprise 35%. This balance reflects the urgency among enterprises to keep pace with rapid AI evolution, as 66% fear the pace of AI innovation is outstripping their ability to adapt.
Competitive Pressures and Market Disruption
Enterprises are acutely aware that AI is not just a tool for optimization—it is a weapon for disruption. More than half (59%) of surveyed IT leaders worry that smaller, more agile competitors who adopt AI faster could displace them. At the same time, 79% believe they have the potential to do the same to larger incumbents, indicating a dynamic market in which size is no longer the key determinant of competitive advantage.
“The data shows both the opportunity and the risk that AI represents,” said Irish. “While excitement is high—73% of CIOs say they feel compelled to adopt AI further—the difference between success and failure will come down to architecture, governance, and control. Enterprises that build a modern data foundation to support critical applications will be those that unlock AI’s full business value.”
Matt McDonough, Senior Vice President of Product at Couchbase, echoed this sentiment: “A modern developer data platform is essential for AI success. Couchbase offers integrated capabilities like vector search and native AI services that allow organizations to build agentic systems at scale. Our multipurpose architecture supports operational, analytical, vector, and mobile data needs in one platform—streamlining AI development and ensuring data integrity, performance, and cost-efficiency.”
About the Study
This survey was commissioned by Couchbase and conducted by Coleman Parkes in April 2025. The study polled 800 senior IT decision-makers, including CIOs, CTOs, and CDOs, from enterprises in the U.S., U.K., Germany, France, Turkey, Japan, India, Singapore, and Australia. All surveyed organizations had 1,000 or more employees.
Learn More
- Download the full report [here]
- Explore key graphics from the study [here]
- Learn how Couchbase empowers agentic system development [here]
- Discover best practices for building modern AI architectures [here]
With AI rapidly becoming the backbone of enterprise competitiveness, the time to act is now. Organizations that seize the opportunity to modernize their data strategy will be better positioned to lead in the AI era.
