OpenAI has submitted a memo to US Congress accusing Chinese AI startup DeepSeek of improperly extracting and using its model outputs through "distillation" techniques to train products like the R1 chatbot, marking the latest escalation in US-China AI competition.
This article analyzes the DeepSeek model distillation incident and proposes a comprehensive multi-layered defense system against distillation attacks, including API-level controls, output watermarking, and architectural protections.
A US federal court has preliminarily ruled that AI image generation tools like Midjourney may have infringed copyright by using artists' works as training data, marking an initial victory for plaintiff artists and igniting fierce debate about balancing AI innovation with intellectual property rights.
OpenAI's revolutionary video generation model Sora faces intense copyright disputes as artists publicly accuse it of plagiarizing original works, sparking heated debates on X platform and highlighting the urgent need for AI training data transparency.
