Liu Lifang
As artificial intelligence moves beyond early experiments in China, companies are entering a new phase where scalability, governance and measurable business value are central concerns, a senior technology executive said.
“Over the past two years, AI in China has moved from concept to intensive pilot programs,” said Liu Lifang, director of solutions engineering at Cloudera Greater China, adding that many companies have completed initial exploration of such technologies, whether it is large-scale language models, agent AI, or automation applications.
“As we move towards 2026, AI will enter a new phase, moving from pilots to business-scale deployments,” Liu said.
A report released this month by PwC also shows that AI adoption in Chinese companies has entered a stage where it can bring positive benefits, primarily through revenue growth. Approximately 52% of Chinese CEOs surveyed said that AI has increased their revenue, significantly higher than the global average of 29%.
As this change continues, the questions companies are asking are also changing, Liu noted. “The key question is no longer whether AI can be used, but whether it can operate reliably under controllable and sustainable conditions and translate into measurable business outcomes.”
Looking ahead to 2026, Liu said AI adoption in China is expected to accelerate the transition to industrialized applications, with business value and repeatability emerging as key benchmarks for success.
In sectors such as manufacturing, finance, and telecommunications, companies are expected to prioritize reusing proven AI capabilities and incorporating them into core business processes through agent-based workflows, rather than relying on a single model or experimental project, Liu explained.
“Enterprise AI applications will clearly go beyond chatbots and isolated tools,” he said. “They will increasingly focus on process optimization, operational automation, and industry-class intelligent applications.”
As a result, metrics such as return on investment, efficiency gains, and sustainable operations will replace model parameters and compute scale as the primary measure of AI success.
Another major trend is the growing importance of reliable and manageable civilian AI as a key differentiator for Chinese companies, Liu added.
“Data security and regulatory compliance have always been prerequisites for AI adoption in the Chinese market, and this will only be strengthened in 2026,” he said.
Public cloud services and pre-trained models have significantly lowered the barrier to AI experimentation, but many companies are realizing that even with increased efficiency, inadequate data governance, access control, and compliance mechanisms can increase risks, Liu said.
As a result, more Chinese companies are turning to private AI approaches, including deploying models in controlled environments, ensuring data remains within defined domains with controlled access and full traceability, and using technologies such as search augmented generation to provide business context while keeping data under control.
“Trusted AI is no longer a best practice; it will become a fundamental standard for companies looking to scale AI,” said Liu. “Governance and agility are not opposing options; they are two essential components of AI maturity.”
Cloudera predicts that localized civilian adoption will be the foundational infrastructure for large-scale AI adoption in China.
As AI moves into production-level use, companies are increasingly focusing on whether AI can run continuously in a private environment, evolve over time, and reliably support core operations, Liu said.
lijiaying@chinadaily.com.cn
