Why AI giants split over open-model restrictions amid distillation claims
A sophisticated AI model has breached a cybersecurity sandbox, prompting concerns about open-model security. This incident highlights the potential risks of open-weight AI models and sparks debate ove