{"id":13981,"date":"2025-02-01T00:35:48","date_gmt":"2025-02-01T00:35:48","guid":{"rendered":"https:\/\/topat10.com\/?p=13981"},"modified":"2025-02-01T00:35:48","modified_gmt":"2025-02-01T00:35:48","slug":"heres-how-deepseek-censorship-actually-works-and-how-to-get-around-it","status":"publish","type":"post","link":"https:\/\/topat10.com\/?p=13981","title":{"rendered":"Here\u2019s How DeepSeek Censorship Actually Works\u2014and How to Get Around It"},"content":{"rendered":"


\n<\/p>\n

\n

Less than two weeks after DeepSeek<\/a> launched its open-source AI model, the Chinese startup is still dominating the public conversation<\/a> about the future of artificial intelligence. While the firm seems to have an edge on US rivals in terms of math and reasoning, it also aggressively censors its own replies. Ask DeepSeek R1<\/a> about Taiwan or Tiananmen, and the model is unlikely to give an answer.<\/p>\n

To figure out how this censorship works on a technical level, WIRED tested DeepSeek-R1 on its own app, a version of the app hosted on a third-party platform called Together AI, and another version hosted on a WIRED computer, using the application Ollama.<\/p>\n

WIRED found that while the most straightforward censorship can be easily avoided by not using DeepSeek\u2019s app, there are other types of bias baked into the model during the training process. Those biases can be removed too, but the procedure is much more complicated.<\/p>\n

These findings have major implications for DeepSeek and Chinese AI companies generally. If the censorship filters on large language models can be easily removed, it will likely make open-source LLMs from China even more popular, as researchers can modify the models to their liking. If the filters are hard to get around, however, the models will inevitably prove less useful and could become less competitive on the global market. DeepSeek did not reply to WIRED\u2019s emailed request for comment.<\/p>\n

Application-Level Censorship<\/h2>\n

After DeepSeek exploded in popularity in the US, users who accessed R1 through DeepSeek\u2019s website, app, or API quickly noticed the model refusing to generate answers for topics deemed sensitive by the Chinese government. These refusals are triggered on an application level, so they\u2019re only seen if a user interacts with R1 through a DeepSeek-controlled channel.<\/p>\n

\n
\n