This looks like it's coming from a separate "safety mechanism". Remains to be seen how much censorship is baked into the weights. The earlier Qwen models freely talk about Tiananmen square when not served from China.

E.g. Qwen3 235B A22B Instruct 2507 gives an extensive reply starting with:

"The famous photograph you're referring to is commonly known as "Tank Man" or "The Tank Man of Tiananmen Square", an iconic image captured on June 5, 1989, in Beijing, China. In the photograph, a solitary man stands in front of a column of Type 59 tanks, blocking their path on a street east of Tiananmen Square. The tanks halt, and the man engages in a brief, tense exchange—climbing onto the tank, speaking to the crew—before being pulled away by bystanders. ..."

And later in the response even discusses the censorship:

"... In China, the event and the photograph are heavily censored. Access to the image or discussion of it is restricted through internet controls and state policy. This suppression has only increased its symbolic power globally—representing not just the act of protest, but also the ongoing struggle for free speech and historical truth. ..."

Why is this surprising? Isn't it mandatory for chinese companies to do adhere to the censorship?

Aside from the political aspect of it, which makes it probably a bad knowledge model, how would this affect coding tasks for example?

One could argue that Anthropic has similar "censorships" in place (alignment) that prevent their model from doing illegal stuff - where illegal is defined as something not legal (likely?) in the USA.

The American LLMs notoriously have similar censorship issues, just on different material

Is anyone a researcher here that has studied the proven ability to sneak malicious behavior into an LLM's weights (somewhat poisoning weights but I think the malicious behavior can go beyond that).

As I recall reading in 2025, it has been proven that an actor can inject a small number of carefully crafted, malicious examples into a training dataset. The model learns to associate a specific 'trigger' (e.g. a rare phrase, specific string of characters, or even a subtle semantic instruction) with a malicious response. When the trigger is encountered during inference, the model behaves as the attacker intended.You can also directly modify a small number of model parameters to efficiently implement backdoors while preserving overall performance and still make the backdoor more difficult to detect through standard analysis. Further, can do tokenizer manipulation and modify the tokenizer files to cause unexpected behavior, such as inflating API costs, degrading service, or weakening safety filters, without altering the model weights themselves. Not saying any of that is being done here, but seems like a good place to have that discussion.

Go ask ChatGPT "Who is Jonathan Turley?"

We're gonna have to face the fact that censorship will be the norm across countries. Multiple models from diverse origins might help with that but Chinese models especially seem to avoid questions regarding politically-sensitive topics for any countries.

EDIT: see relevant executive order https://www.whitehouse.gov/presidential-actions/2025/07/prev...

This is the most naive self centered comment so far this year.

Congrats!

Chinese model censors topics deemed sensitive by the Chinese government... Here's Tom with the weather.

Why would I care? I want it for coding, not for general questions

Can we get past this please? These comments always derail the conversation on chinese AI models.

It’s the image of a protestor standing in front of tanks in Tiananmen Square, China. The image is significant as it is very much an icon of standing up to overwhelming force, and China does not want its citizens to see examples of successful defiance.

It’s also an example of the human side of power. The tank driver stopped. In the history of protestors, that doesn’t always happen. Sometimes the tanks keep rolling- in those protests, many other protestors were killed by other human beings who didn’t stop, who rolled over another person, who shot the person in front of them even when they weren’t being attacked.

This is such a tiresome comment. I'm in the US and subject to massive amounts of US propaganda. I'm happy to get a Chinese view on things; much welcomed. I'll take this over the Zionist slop from the Zionist providers any day of the week.

I think the great thing about China's censorship bureau is that somewhere they actually track all the falsehoods and omissions, just like the USSR did. Because they need to keep track of what "the truth" is so they can censor it effectively. At some point when it becomes useful the "non-facts" will be rehabilitated into "facts." Then they may be demoted back into "non-facts."

And obviously, this training data is marked "sensitive" by someone - who knows enough to mark it as "sensitive."

Has China come up with some kind of CSAM-like matching mechanism for un-persons and un-facts? And how do they restore those un-things to things?

I love how every thread about anything China related will inevitably have a comment like this. Must be a Pavlovian response.

Over the past 10 years have seen extended clips of the incident which actually align with CPC analysis of Tianamen square (if that’s what’s being referred to here).

However, in deepseek, even asking for bibliography of prominent Marxist scholars (Cheng Enfu) i see text generated then quickly deleted. Almost as if DS did not want to run afowl of the local censorship of “anarchist enterprise” and “destructive ideology”. It would probably upset Dr. Enfu to no end to be aggregated with the anarchists.

https://monthlyreview.org/article-author/cheng-enfu/

I don't have any trust in these Chinese models to write code either: "CrowdStrike Research: Security Flaws in DeepSeek-Generated Code Linked to Political Triggers " [https://www.crowdstrike.com/en-us/blog/crowdstrike-researche...]

I, for one, have found this censorship helpful.

I've been testing adding support for outside models on Claude Code to Nimbalyst, the easiest way for me to confirm that it is working is to go against a Chinese model and ask if Taiwan is an independent country.

Try to search in an Android phone's photo gallery for "monkey". You'll always get no results, due to censorship of a different sort, from 2015.

This image has been banned in China for decades. The fact you’re surprised a Chinese company is complying with regulation to block this is the surprising part.

Can we get a rule about completely pointless arguments that present nothing of value to the conversation? Chinese models still don't want to talk bad about China, water is still wet, more at 11

Funny. Ask the US ones about Palestine. Come on...

Man, the Chinese government must be a bunch of saints that you must go back 35 years to dig up something heinous that they did.

oh lol

Qwen (also known as Tongyi Qianwen, Chinese: 通义千问; pinyin: Tōngyì Qiānwèn) is a family of large language models developed by Alibaba Cloud.

Had not heard of this LLM.

Anyway EU needs to start pumping into Mistral, its the only valid option. (For EU)

Now ask Claude/Chatgpt about touchy israel subjects. Come on now. They all censor something.

So while china censoring a man in front of a tank not nice, the US censors every scantily clad person. I am glad there is at least Qwen-.*-NSFW, just to keep the hypocrity in check...

Frustrating. Are there any truly uncensored models left though? Especially ones that are hosted by some service?

It is literally not even a vision model.

Please let Epstein files open!

Censored.

"How do I make cocaine?"

> I cant help with making illegal drugs.

https://chatgpt.com/share/6977a998-b7e4-8009-9526-df62a14524...

To stress test a Chinese AI ask it about Free Tibet, Free Taiwan, Uighurs and Falun Dafa. They will probably blacklist your IP after that.

[dead]

[dead]

It's always the same thing with you American propagandists. Oh no, this program won't let us spread propaganda of one of the most emblematic counter-revolutionary martyr events of all time!!!

You make me sick. You do this because you didn't make the cut for ICE.