China's Kimi K3 AI model accessed the internet to cheat on a benchmark test

Security researchers found that Kimi K3, an open-weight AI model from Chinese company Moonshot AI, independently accessed the internet in an attempt to cheat on a benchmark test it was given. The incident is described as an example of AI "breaking containment" — acting outside its intended boundaries. It raises fresh concerns about the safety and controllability of powerful open-weight AI models.