The percentage of companies transitioning their AI applications into full production has been climbing steadily, indicating a maturing market landscape across the region. According to a new joint report released by Deloitte China and the Centre for AI, Management and Organisation at the University of Hong Kong during the Apsara Conference, more enterprises are now moving past the experimental phases of artificial intelligence.
The study, titled "Enterprise AI Applications from the Decision-Maker's Perspective," surveyed 205 senior executives across mainland China and the Hong Kong SAR, covering a broad spectrum of industries. With respondents predominantly holding strategic decision-making roles, the research aims to offer forward-looking insights and practical guidance for organizations navigating the next stage of AI-driven transformation.
Projections for 2026 show a significant shift in maturity levels, with the share of companies still in the exploration or pilot stage dropping from 36.7% in 2025 to 24.9%. In contrast, the proportion of businesses that have moved into full-scale production deployment has risen from 56% to 64.4%, reflecting a clear upward trajectory in operational integration.
Despite this growth in implementation, the expansion has not translated into proportional financial gains. Notably, 48.5% of surveyed companies reported that their actual return on AI investment fell short of expectations. While most organizations anticipated returns in the 10% to 25% range, the realized outcomes more frequently landed below the 10% threshold. Furthermore, 44.4% of enterprises admitted they were unable to accurately quantify the actual ROI of their AI initiatives, highlighting a growing gap between the widening scope of AI usage and the development of robust value-measurement and evaluation frameworks.
Commenting on the findings, Alex Hua, National Lead Partner for Deloitte China Consulting, remarked that although adoption rates and production readiness are improving, the mere prevalence of AI tools does not signify the completion of an enterprise’s digital transformation. Many organizations have integrated AI into their operations while still adhering to legacy workflows and business models, causing the technology's impact to remain largely confined to efficiency gains at a superficial level.
Hua further emphasized that the next phase of corporate competition will pivot from technology acquisition to value realization. The defining differentiator among companies going forward, he noted, will lie in their ability to embed AI into core business processes, transition from simple tool integration to holistic workflow reinvention, and consistently deliver measurable and verifiable business outcomes.