# Chinese AI models often echo Party line, study finds

> Aleph Alpha tested 967 political prompts and saw Chinese models give biased answers, while even NVIDIA’s Nemotron showed some Party‑aligned data.

Oossa · 2026-09-29 · https://oossa.com/en/chinese-ai-models-often-echo-party-line-study-finds

Aleph Alpha evaluated 12 Chinese open‑weight LLMs and found they answered only 17‑41 % of politically sensitive prompts in a balanced way. The test also flagged NVIDIA’s Nemotron Cascade 2 on 17 % of prompts because about 3.5 k of its 9.3 M training rows contain Chinese Communist Party talking points. The company says future models will screen for such content, add targeted alignment data, and use benchmarks like this one.

## The facts

- 967 political prompts were used in the benchmark
- 3.5 k of Nemotron’s 9.3 M chat rows carried CCP‑aligned text

## Why it matters

Understanding political bias in AI helps users and regulators decide how safe and neutral these models are for everyday use.

## Sources & references

1. [Training on the Party Line: Chinese Political Influence on LLMs in China and the World](https://www.aleph-alpha.com/en/blog/training-on-the-party-line/) – Aleph Alpha, 2026-09-29

Last updated: 2026-09-29