Hong Kong businesses are adopting generative artificial intelligence much slower than mainland Chinese enterprises, hindered by legacy systems and a risk-averse culture. Only 10 per cent of Hong Kong firms surveyed by Accenture have integrated AI across multiple business functions, compared to roughly 21 per cent on the mainland.
The digital transition in Hong Kong faces distinct hurdles that separate local businesses from their counterparts across the border. While mainland enterprises move quickly to embed AI directly into daily operations, firms in the city remain weighed down by older technology infrastructure and a corporate outlook that avoids operational risks, according to Robert Hah, Hong Kong office lead at Accenture Greater China, in an interview with the South China Morning Post on Tuesday.
Entrenched Legacy Systems and Cautious Corporate Culture Slow Local Uptake
The contrast in adoption rates highlights a widening gap in commercial technology integration within the region. Research from the consultancy indicates that just 10 per cent of surveyed Hong Kong companies had integrated generative AI across multiple business functions, as reported on September 29, 2026. In contrast, roughly 21 per cent of mainland enterprises have achieved similar levels of integration.
The disparity reflects a digital transition in the city that is being slowed by older technology foundations and a corporate culture that avoids risk, the consultancy executive indicated.
Mainland Developer Ecosystem and Falling Model Costs Fuel Rapid Expansion
Across the border, the commercial AI market benefits from a flourishing domestic developer ecosystem and declining model costs. Mainland China’s rapid adoption of enterprise AI has been fuelled by falling model costs and a flourishing domestic developer ecosystem led by firms such as DeepSeek and Alibaba Group Holding, owner of the South China Morning Post.
Major technology firms including Ant Group, Tencent Holdings, Alibaba, and Baidu rushed to roll out enterprise-focused AI agents at this year’s World Artificial Intelligence Conference in Shanghai, the country’s top AI event. Tech giants including Ant Group, Tencent Holdings, Alibaba and Baidu rushed to roll out enterprise-focused AI agents at this year’s World Artificial Intelligence Conference in Shanghai – the country’s top AI event – signalling a shift in the industry battleground from building foundation models to embedding AI directly into corporate workflows.
Broader Challenges in Translating AI Ambition Into Measurable Value
The challenge of turning artificial intelligence initiatives into bottom-line returns extends beyond simple regional comparisons. A separate report by Deloitte China and the University of Hong Kong (HKU) highlights a similar challenge across China, where companies are struggling to translate AI ambition into measurable outcomes. A Deloitte China and HKU AI Adoption Index 2026 survey found that 56% of organizations are in limited AI implementation.
This index, based on a survey of over 100 C-suite leaders, found that while confidence in AI’s long-term potential is high, scalable results remain elusive. The same survey indicated that 23% of organizations report measurable financial impact from AI, with only 4% being transformational. The primary barriers are identified as organizational issues such as siloed departments, cultural resistance, and poor data management, rather than technological limitations.
Strategic Integration Versus Shallow Tool Application
Accenture’s research also indicates that while Chinese companies are accelerating AI adoption, only about 9% are realizing significant value in terms of productivity, revenue, or profitability. Many implementations remain shallow, with companies stuck in the tool application phase rather than achieving strategic integration.
Despite these challenges, Chinese firms continue to invest heavily in digital transformation, with their Digital Transformation Index rising. Examples like Wuxi Lead Intelligent Equipment and Midea Group demonstrate successful AI deployment, achieving high defect detection accuracy and significant cost savings, respectively.