In November 2025, Deloitte China, Ant International, and the Global Finance & Technology Network published the AI Inclusion Report. Citing the Stanford 2025 AI Index, the report notes that about 78% of enterprises worldwide had integrated some form of AI technology into core operations in 2024, up from 32% in 2020. This is an overall global enterprise adoption measure, not the adoption rate of Taiwanese SMEs, and it must not be combined into the same denominator as the Ministry of Economic Affairs (MOEA) statement that "91% of assisted enterprises are SMEs."
What the report actually highlights is the capability gap facing SMEs, not guaranteed outcomes. Citing OECD data, Deloitte notes that only 8% of SMEs have reached a transformative level of digital maturity that prioritizes digital innovation and advanced tools; infrastructure costs, talent and skills, and the ability to access policy and ecosystem support still constrain AI adoption. Rising global adoption does not mean every SME already has the necessary data, process, and governance conditions.
The quantified revenue outcome claimed in the original draft could not be found in this Deloitte report or its official introduction page, and no sample, period, comparison baseline, or calculation method was provided, so it has been removed in this revision. The report discusses five principles, lowering tool barriers, building AI literacy, responsible use, public-private collaboration, and talent capabilities, and should not be rewritten as a fixed revenue outcome or a survival crisis.
A more practical starting point for SMEs is to choose one high-frequency, measurable task that is reversible if it fails, such as reconciliation, cash flow forecasting, customer service triage, or organizing quotation data. Before adoption, record labor hours, wait times, error rates, and manual rework; then confirm data sources, system permissions, human approval points, exception ownership, and stop conditions, and expand only after targets are met.
Government resources can lower the cost of diagnostics, training, or pilots, but program years, eligibility, application status, and grant conditions change, and funding does not equal adoption results. From the perspective of "The CIO for Your Business Operations," Unima recommends breaking AI adoption into four acceptance areas: use-case value, data foundation, team capability, and governance accountability. This is a practical recommendation compiled from the Deloitte report and does not represent specific client results or revenue commitments.
First-party sources: Deloitte China, "The AI Inclusion is Here: How Can SMEs Leverage It to Break Through?", https://www.deloitte.com/cn/en/Industries/tmt/perspectives/ai-inclusion-report.html; Deloitte China, Ant International, Global Finance & Technology Network, "AI Inclusion Report – Inclusive Finance Powered by AI," 2025-11, https://www.deloitte.com/content/dam/assets-zone1/cn/en/docs/industries/technology-media-telecommunications/2025/ai-inclusion-report-en-251113.pdf. This article was fact-checked on 2026-08-25; all percentages are limited to the global enterprise or SME maturity measures as defined by the sources.