Irvine, Calif. – November 04, 2026 -- Only 22% of organizations surveyed by Gartner have scaled artificial intelligence across multiple business units or adopted an AI-first approach, even though 85% of surveyed functional leaders plan to increase AI spending in 2026, according to enterprise technology provider Trace3. The company says the gap becomes expensive when tools reach departments faster than they can coordinate purchasing, prepare data, set safeguards and measure returns.
Deloitte data shows tool access outpacing process redesign
Deloitte's 2026 State of AI in the Enterprise report found that workforce access to sanctioned AI tools rose from under 40% to around 60% within a year. Only 30% of surveyed organizations were redesigning key processes around AI.
Kevin Rogers, Practice Director of Data and Analytics at Trace3, described companies that bought two separate instances of the same type of cloud data platform. One was for security vulnerability analysis and the other for enterprise data and analytics. Neither buyer recognized the chance to consolidate, and the companies lost economies of scale.
Duplication also occurs in data work, Rogers said. Separate teams may each prepare the same enterprise data for their own AI use cases, repeating effort across platforms and data pipelines.
Trace3 recommends a shared inventory and center of excellence to expose overlap
Rogers advises cataloging existing tools, planned pilots and data resources before approving new spending. He says pockets of AI expertise scattered across a company should be tied together through a center of excellence.
Value and feasibility scoring should rank AI projects before funding
Rogers recommends ranking use cases by business value and technical feasibility. High-value projects with manageable technical requirements can deliver early evidence before companies commit more resources.
Financial analysis should also capture costs often missed in planning, including employee time on manual tasks, model usage and ongoing operating expenses. Rogers said companies should measure a process's cost before AI and after implementation.
Usage-based AI models complicate that comparison. Unmonitored consumption can reduce or erase an initiative's intended savings. Rogers said organizations should track usage and set guardrails on token consumption, particularly with expensive models, so operating costs can be set against the value each initiative delivers.
Fragmented data architecture can stall AI programs for years
Preparing enterprise data for AI can expose problems that existing systems concealed. Information may be scattered across platforms, inconsistently structured, or governed by access permissions that do not match how AI applications will use it.
Data architecture and modeling require specialized expertise, Rogers said, and internal teams often maintain existing systems while also building new AI capabilities. Some organizations spend years working through these foundational issues.
Trace3 assesses client data environments, identifies shared infrastructure and supports the technical work of preparing information for AI. Rogers said the priority is proprietary data that competitors lack, which must be available, centralized, governed and secure, with enough context for AI systems to interpret it.