The impact of large language models on the job market in China is becoming increasingly apparent, as the recruitment scale for occupations most affected by these models declined between 2022 and 2024.

According to a report jointly released by the National School of Development at Peking University and recruitment platform Zhaopin last year, the Work Activity-based LLM Exposure Index reached 0.89 in 2024 for the editing and translation category, and 0.68 for the category covering human resources, administration, finance and legal professions — an increase of 0.09 and 0.04, respectively, compared to 2022.

The report also indicates a negative correlation between the LLM Exposure Index and job postings: the higher the index score, the greater the decline in job postings in China between 2022 and 2024.

For example, in fields with a higher index score, such as sales and business development, the proportion of job postings in the first half of 2024 decreased by more than five percentage points compared to 2022.

Similarly, job shares in software and hardware development, operations, testing, human resources, administration, finance and legal professions declined by over one percentage point during the same period.

The report suggests that this reduction in job postings may have also been influenced by multiple factors, including industrial restructuring and advancements in large AI model technology.