Anthropic and OpenAI seek narrower Australian copyright limits for AI training

Anthropic and OpenAI asked an Australian parliamentary inquiry to relax copyright rules that require AI developers to obtain permission before training on protected local works. Anthropic proposed narrow conditional approval for model training, potentially tied to investment or support for Australian creators. OpenAI requested a framework allowing training on publicly available information while offering ways for rights holders to work with AI companies. Both linked copyright policy to planned Australian data centers. The government ruled out a text-and-data-mining exception in October 2025, while creative-industry advocates maintain that developers should pay to use copyrighted material.