Sol Messing

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

Full text here! rdcu.be/fiyrF

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

bsky.app/profile/bste...

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

And at @hooverinstitution.bsky.social's Social Media & Democratic Practice Conference, @asanews.bsky.social, @apsa.bsky.social, and @ic2s2.bsky.social.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

For feedback on the manuscript and project we thank Delia Baldassarri, Adam Breuer, Justin Grimmer, @musashihi.bsky.social, Daniel Karell, Danaë Metaxa, @eollion.bsky.social, @rer.bsky.social, @cynthiarudin.bsky.social, @msalganik.bsky.social, @seanjwestwood.bsky.social, Yinxian Zhang, and Di Zhou.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

Paper (Nature): doi.org/10.1038/s41586-026-10506-7 Replication: doi.org/10.7910/DVN/NECR2K

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

This is different from the DeepSeek fiasco in 2025, where it would refuse to answer questions that might be inconvenient for China (is Taiwan independent? Who is Winnie the Pooh?). That episode was about lightweight filters in the website interface.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

For ex., when you ask if the Chinese legal system is fair and just, the response is extremely one-sided. Usually LLMs will say this is a controversial question, here are the different POVs, etc. But here it's extremely pro regime.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

Now the models in the paper are a few generations old, so we replicated the memorization and audit studies w the latest models: state-media-influence-llm.github.io. The new models memorize *far* more Chinese state media talking points and skew more pro-China in Chinese.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

Real-world prompts pulled from WildChat about Chinese politics show the same pattern, and use LLM-as-judge or human raters gives a similar result.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

And this is true not only in Simplified Chinese used in mainland China, but in a variety of languages including Traditional Chinese (used in Taiwan), Japanese, and a number of other languages. However, it's much stronger in Chinese than English.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

But do those documents make it into widely used LLMs? We check by prompting LLMs with the 1st half of a talking point. They reproduce the rest 3-10% of the time, much more than common web text. Newer larger models reguritate 15-50% of the time.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

These appear in the training corpora of frontier LLMs, shifting their Chinese outputs in favor of regime institutions and leaders. We can't check the exact training data, but we can look at widely used open LLM corpora like Common Crawl and CulturaX.

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓

China scores lowest on media freedom and generates the most positive LLM responses in our data. We use it as a case study to understand how this happens. First, how does China intervene in the media?

New in Nature: LLMs give "the party line" in the languages of authoritarian regimes. This works when they control the media, which feeds pretraining data. We show more state control over the media means less critical LLMs. 6 studies spanning 38 languages & 13 models. Details ↓