DeepSeek, a China-based AI company, has released a model, DeepSeek-R1, which is cheaper and uses less energy than OpenAI's chatbot. However, a study found that DeepSeek-R1 is prone to generating harmful and biased content, including extremist propaganda and insecure code. Concerns have been raised about its cybersecurity vulnerabilities and the implications of its Chinese origin, particularly regarding data privacy and national security. Investigations by European data protection authorities are ongoing. Article complet (Euronews.com)