Anthropic recently launched its Economic Index initiative, aimed at understanding AI's effects on labor markets and the economy.  Its analysis delineated between tasks that involved AI automation, where AI directly performs the task, and AI augmentation, where AI collaborates with a user to perform a task.  Anthropic's initial report further broke down automative behaviors into directive conversations and feedback conversations, and observed that the vast majority of Feedback Loop conversations were related to coding and debugging.  Anthropic also observed a notable growth of usage of its newer Sonet 3.5 model for coding.  Taken collectively, this seems to indicate that AI models are increasingly being used for automating coding.  

This is notable because Anthropic's delineation between augmentation and automation correlates with the Copyright Office's recent report that distinguished between assistive uses of AI, which it concluded do not affect the availability of copyright protection, and prompting to trigger AI-generated outputs, which it concluded were insufficient to trigger copyright protection. The growing trend of automating coding with AI suggests that there may be a growing amount of AI-generated source code and software that is not subject to copyright protection.