Enterprise AI Stalls in Proof-of-Concept Trap as Nearly 60% of Global Firms Struggle to Scale: Citi Report

Despite corporate boards allocating unprecedented capital to artificial intelligence over the past two years, the enterprise sector is grappling with an execution bottleneck. According to a new report by Citi Institute, nearly 60% of surveyed organizations remain stuck in exploration or pilot stages, with only about 40% managing to transition solutions into production or enterprise-wide scaling.

The findings indicate that the next phase of corporate AI adoption will hinge on whether organizational leadership can bridge the gap between proof-of-concept experiments and live operational workflows capable of delivering measurable business returns.

The Regional Adoption Divide The transition from pilot sandboxes to operational execution is progressing unevenly across global business corridors:

• North American Frontrunners: North America leads the adoption curve, with 77% of surveyed organizations actively running AI in live production or scaling it across core departments.

• Asia-Pacific Sandbox Stagnation: In contrast, firms across the Asia-Pacific region remain heavily focused on early capability development, exploratory testing, and small-scale pilot trials.

• Europe's Intermediate Balancing Act: European organizations occupy the middle ground, negotiating stricter regulatory compliance frameworks and data residency mandates while gradually transitioning models into active business units.

The Scaling Dividend: 15%+ Measurable Impact For the 40% of enterprises that have successfully navigated technical and cultural hurdles to deploy AI into production, the business upside is clear:

• Substantial Value Creation: Sixty percent of organizations operating in production and scaling phases report transformational value creation exceeding 15%, driven by structural cost reductions, enhanced workforce productivity, and direct revenue growth.

• Strategic Intent: Business model transformation has emerged as the primary motivator for corporate AI spending, cited by 34% of respondents, followed by organizational capability building at 32% and targeted key performance indicator (KPI) improvements at 25%.

"AI's biggest opportunity lies in redesigning business models, pointing to the potential for companies to use the technology to reshape processes, products and revenue streams rather than focus only on automating existing tasks." — Citi Institute Report

The Operational Pivot: From Routine Tasks to Agentic Systems While basic task automation remains the baseline entry point across the corporate landscape, organizations deploying substantial technology budgets are pivoting toward autonomous agents capable of navigating complex, multi-layered business processes.

The report highlights that long-term enterprise differentiation will not stem from access to commercial frontier models alone. Instead, execution relies on proprietary domain data, deep in-house engineering expertise, and seamless integration into core enterprise operations. Furthermore, companies advancing into production are adopting noticeably shorter payback horizons, indicating that enterprise leadership is treating artificial intelligence as a mandatory operational capability rather than an open-ended research exercise.