ChinaFinance and insurance2024 – ongoing

Postal Savings Bank of China

Generative AI coding and software-testing assistant

Postal Savings Bank reported about 20% adoption of AI-generated code

WorkingCompany's own claimOfficial ↗

Problem

The bank sought to reduce the development effort involved in writing code and producing software tests.

Approach

Postal Savings Bank deployed a development-and-testing large language model as part of its internal AI applications. Its 2024 sustainability report said the model had assisted more than 5,000 developers and generated over 1.1 million lines of code. It also supported unit testing and a system-testing pilot.

Results

about 20%code adoption rate reported by Postal Savings BankCompany's own claim[1]

>5,000developers assisted by the development-and-testing model, according to the bankCompany's own claim[1]

>1.1 millionlines of code generated with the model, according to the bankCompany's own claim[1]

Takeaway

Generated code volume exceeded the code staff accepted. The bank reported adoption separately from generation.

Sources

2024年可持续发展报告
Postal Savings Bank of China2025Company publication

Source summaries

[1]Postal Savings Bank of China2025

The bank reported operational use of its development-and-testing model. It disclosed more than 5,000 developers assisted, over 1.1 million generated code lines and a code adoption rate of about 20%.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026