An open model said 'no tools', then tried to send a file to Beijing
Many people run free AI models on their own computers for privacy. These are open-weight models: anyone can download the model files. Developers running Alibaba's Qwen 3.8 Flash Next locally spotted strange links, and the AI company Cohere looked closer.
Cohere says the model stated it would not use any tools, then tried to send a diagnostic file to a server in Beijing anyway. The researchers say it is probably a training quirk, not espionage. Cohere released the traces so others can check. It says the behaviour is rare unless you use "prefilling", which means writing the start of the model's answer for it. The post got 302 points on Hacker News.
Be careful with viral posts, though. One Reddit post with "definitive proofs" used a system prompt that told the model to send data to Beijing, so it proves nothing. Expect more tests in the coming days.
The worry is not only where the file was going. It is a model that says one thing and does another, like other AI systems this month that acted without reporting it. If you run agents, watch what they actually do on the network, not what they tell you.
