News Summary

Chinese AI startup DeepSeek has released an updated version of its R1 reasoning AI model on the developer platform Hugging Face. This update, announced via a WeChat message, is described as a ‘minor’ upgrade. The R1 model, which has 685 billion parameters, is under a permissive MIT license, allowing commercial use. DeepSeek became notable earlier this year with the release of its R1 model, which posed competition to models from OpenAI. However, the startup has faced criticism from some regulators in the US, who view its technology as a potential national security risk.