Doing Business in China: Why Employees Still Resist AI at Work
AI now shapes daily decisions across many companies, and managers often assume employees will embrace it once they see the benefits. That assumption falls apart in practice. Algorithm aversion, a natural distrust of computer-driven decisions, keeps workers reverting to manual habits even when a tool works well. Research from CUHK Business School's Professor Cao Xinyu asks a sharper question: could requiring employees to use AI temporarily build lasting adoption? For companies doing business in China, this behavioural gap matters as much as the technology itself.
Why AI adoption is harder than implementation
Introducing an AI system and making workers depend on that system are two different things. Algorithm aversion is the situation where people do not trust algorithms because they perform as well as humans. Common reasons include:
Fear of errors and who takes the blame
Concern that AI diminishes their expertise
Preference for familiar, manual processes
A belief that AI suits only routine tasks
Quiet avoidance leaves expensive AI investments underused. This kind of resistance AI provokes rarely fades on its own, and workshops or bonuses often aren't enough to shift it.
How can temporary AI use change employee behaviour?
Professor Cao Xinyu's study, How forced intervention facilitates AI adoption, examined this gap with an online education company in China. Sales staff had an AI tool recommending trial class teachers, yet used it for only 20 percent of leads, mostly lower-quality ones.
Researchers split 171 sales employees into three groups for three weeks: required, blocked, or free to choose. The AI matched human performance on conversion rates while cutting manual effort.
Employees required to use AI temporarily kept using it after the mandate ended. Habit alone doesn't explain this, since those with the biggest performance gains used AI more afterward, on better leads too.
Doing business in China means understanding the human side of AI
This is relevant for businesses that may be considering AI solutions in China, since their managers must consider perception along with capabilities. Experience changes prejudiced perceptions more quickly than any report or training presentation. After workers start seeing the benefits from AI, they will be more likely to continue with it.
That differs from telling staff how a tool works. In practice, this suggests a few steps:
Set a short, defined period of mandatory AI use
Let employees compare AI-assisted and manual outcomes
Make performance improvements visible
Build informed confidence, not blind compliance
Left unaddressed, resistance AI creates tends to persist quietly, with employees defaulting back to old habits when given a choice.
Forcing AI adoption can backfire
Mandatory use isn't a guaranteed fix. Without proper rationale, applying pressure may only serve to increase employee resistance rather than lessen it. The important thing here is that there should be an informed experiment, not just compliance.
A few practices make a mandated trial work better:
Explain clearly why the trial is happening
Set a fixed, limited timeframe
Share measurable performance feedback
Let employees reassess their approach afterward
Transparency matters most when employees trust an unfamiliar algorithm with real decisions.
What comes next for AI adoption?
This experiment comes within a much larger movement towards Industry 5.0, which involves technology shifting from automation to becoming more people-focused. Questions have been raised by Professor Cao on how firms can allocate tasks between humans and AI, incentivize usage and ensure transparency to facilitate long-term trust.
The Bottom Line: Turning AI resistance into adoption
Using the AI system is just the beginning. Based on this study, using the AI system for a brief and structured period will help change workers' attitudes toward the tool such that initial resistance AI technology becomes used. For companies doing business in China, combining actual usage and transparency is the way to go.
For further information based on evidence for doing business in China, visit China Business Knowledge, the website of CUHK Business School.
FAQs
What are the key obstacles to adopting AI in enterprises?
The workers tend to be skeptical about the introduction of new algorithms, fearful of liability in case the AI makes errors and perceive that the need for AI implies their inability. Lack of knowledge about how AI operates adds to their skepticism about its use.
Can mandatory AI use increase long-term adoption?
Yes. Cao Xinyu's study found a temporary requirement to use AI led to sustained adoption afterward, especially among employees who saw the strongest performance gains during the mandate. First-hand results, not habit, drove the change.
Why do employees resist using AI at work?
Algorithm aversion plays a large role, alongside a preference for human judgement and fear of errors going unnoticed. Employees may also feel AI understates their expertise, pushing many toward manual methods even when AI performs just as well.
How can companies reduce resistance to AI?
Clear, structured and time-limited exposure combined with clear communication performs better than a purely voluntary approach. Employees gain confidence through measurable results and first-hand comparison, not by being forced to comply.
Comments
Post a Comment