China is reportedly preparing export controls on artificial intelligence, with national security given as the reason.
That is close to everything known so far. Reuters reported the plan, and the one detail beyond the headline is a reassurance: people will still be able to use DeepSeek, the Chinese AI company, and API access will remain. An API is the connection developers use to plug a model into their own software.
What the reporting leaves out matters as much. It does not say what the controls would cover, when they would start, or how they would be enforced. It names no affected companies or technologies. Until the government publishes something, "export controls" could mean anything from a licensing step to a narrow ban.
The direction of travel is the notable part. Export controls on AI have mostly been a US tool aimed at China, most visibly the restrictions on advanced AI chips sold there since 2022. A Chinese version would reverse the arrangement. China has used the mechanism before, restricting exports of materials such as gallium and germanium in 2023, but applying it to AI would be a different kind of target.
The DeepSeek promise is also narrow. It covers use and API access, and nothing more specific. Whether any rules would reach the model files that developers elsewhere download and build on is a question the reporting does not answer.