Meta Platforms Inc. announced in July that it would release Llama 3.1, the “first frontier-level open source AI model” and one of the world’s largest publicly available pre-trained large language models (LLM), making the LLM’s code and underlying architecture freely available.
The release of a pre-trained open source model creates a curious dynamic in the question of fair use: Is copyright infringement more defensible under the fair use doctrine when considering an open source model or a closed source model? And did that play a role in Meta’s decision to make its LLM source code publicly available?
Courts handling copyright ...