Taobao Instant Commerce, China’s largest online delivery service, handles about 40 million orders per day. Couriers frequently call customers to coordinate the final handoff, which creates a barrier for workers who are deaf or hard of hearing.
A collaborative study by researchers at the University of Toronto, the University of California, Santa Barbara, and Zhejiang University found that deaf and hard-of-hearing couriers earned roughly 10% less per hour than their colleagues. The study also found these couriers received more negative customer ratings and were more likely to make late deliveries.
In response, three engineers at Taobao developed a simple accessibility feature in their off hours: couriers can select preset messages (or write custom ones), and an AI system places a call to the customer and reads the message aloud. The feature was designed to remove the need for live phone coordination that disadvantaged hearing-impaired workers.
Researchers estimated the tool closed about one-third of the hourly pay gap. Late deliveries and negative ratings also declined after the feature’s introduction, with the largest benefits accruing to couriers with more severe hearing loss.
The feature fits into a broader accessibility push in China, where businesses are required to reserve at least 1.5% of job opportunities for people with disabilities. After Taobao rolled out the AI calling feature, rival Meituan implemented a similar solution.
The case highlights how relatively small software changes can produce measurable improvements in workplace equity and may prompt competitive adoption across the industry.
Why this matters
- Economic impact: deaf and hard-of-hearing couriers earned about 10% less per hour.
- Measurable improvement: the AI calling tool is estimated to have closed roughly one-third of that gap.
- Practical solution: AI reads preset or custom messages to customers in place of live phone calls.
- Policy context: Chinese firms must reserve at least 1.5% of positions for people with disabilities.
This example shows technology can reduce practical barriers that translate into real wage and performance differences for disabled workers.
