Worldwide spending on artificial intelligence (AI) is projected to reach $2.52 trillion by 2026, reflecting a 44 percent increase year-on-year, according to research firm Gartner. Despite this surge in investment, nearly two-thirds of organizations have yet to scale AI across their enterprises. While 88 percent of companies report using AI in at least one business function, the challenge now lies in translating those efforts into scalable value.
Business leaders generally grasp what is needed for successful AI scaling: workflows must decentralize, AI ownership should be closer to decision-making business units, and integration into daily operations is critical. Clean and reliable data is also essential, as noted in multiple AI strategy documents and consultant frameworks. However, a recent survey by Alteryx, which included responses from 1,400 business and IT leaders globally, highlights a concerning statistic: fewer than one in four AI pilots successfully transition to full-scale production.
The foundational requirements for scaling AI—decentralizing workflows, embedding AI into core systems, and ensuring data quality—are well understood. For instance, by 2028, responsibility for AI workflows is expected to shift to individual lines of business, with 33 percent of workflows decentralized. Yet, even as organizations improve data quality and AI integration, many find that progress does not necessarily lead to measurable impact. The key issue remains: when AI operates across decentralized units, who ensures that central governance controls are effectively applied at the point of decision-making?
Organizations that have successfully scaled their AI initiatives—approximately 23 percent, according to Alteryx—share a distinct profile. These organizations report high data maturity, robust governance adherence, and a commitment to transparency. They do not just have these capabilities; they govern them consistently in practice. Notably, in Singapore, 60 percent of respondents indicated that centralized data governance is a missing capability, higher than the global average of 53 percent. This disparity underscores that while governance intent is widespread, effective execution remains elusive.
Many organizations have established AI governance frameworks, yet the challenge lies in ensuring these frameworks extend beyond central oversight. Effective governance must be enforced at the point where AI is utilized, rather than solely defined in policy documents. A policy that outlines approved data sources and accountability structures is a necessary foundation, but it does not guarantee compliance when decisions are made.
Failures in scaling AI are predominantly operational rather than technical. Governance that exists only on paper cannot sustain decentralized AI efforts. The case of FWD Insurance in Hong Kong exemplifies this challenge. As a rapidly growing insurer operating across ten markets, FWD faced increasing reporting demands and regulatory complexities. Their finance and actuarial teams relied on manual workflows, which extended month-end reporting and scenario testing by several days.
The real hurdle for FWD was translating governance intent into operational practices. By deploying Alteryx and integrating automated workflows into its finance and actuarial functions, FWD achieved remarkable results: a 95 percent reduction in reporting time, an 80 percent decrease in data preparation errors, and a monthly savings of ten days in reporting. Alteryx One played a critical role by embedding governance as a core function, enforcing controls at the point of execution.
FWD’s Director of Finance Transformation described the initiative as creating a smarter, more agile finance function, effectively operationalizing governance through clean data, auditable workflows, and accountability where decisions are made.
In a broader context, Singapore recently became the first government globally to introduce a governance framework specifically for agentic AI, signaling a significant step toward responsible AI deployment. However, compliance remains voluntary, leaving organizations to interpret and apply governance as they see fit. The framework exists, but the level of seriousness with which each organization adopts it varies.
For organizations aiming to scale AI successfully, decentralized workflows, tight integration into daily operations, and high-quality governed data are necessary conditions. Nonetheless, these conditions alone do not guarantee positive outcomes. The real catalyst for measurable business impact—lower costs, faster decisions, and enhanced performance—lies in governance that functions as an enforcement mechanism rather than merely an aspiration.
See also
AI Transforms Health Care Workflows, Elevating Patient Care and Outcomes
Tamil Nadu’s Anbil Mahesh Seeks Exemption for In-Service Teachers from TET Requirements
Top AI Note-Taking Apps of 2026: Boost Productivity with 95% Accurate Transcriptions
Machine Learning Tools Transform Materials Discovery, Enhancing Semiconductor Efficiency
