Meta used copyright to protect its AI model, but argues against the law for everyone else.

The interest of Meta, the company behind Facebook, Instagram and Whatsapp, in artificial intelligence is not exactly new. In an attempt to achieve OpenAI's fame, the company has invested with peculiar constancy in various projects dedicated to the artificial generation environment over the past year. 

Recently, Meta stated that since the content used to train its AI models is freely available on the Internet, the use made by the company falls under the definition of 'fair use'. This law regulates "the ability to use copyrighted material for purposes of information, criticism or teaching without seeking written permission from the rights holder."

Ironically, Meta's language model was made publicly available during 2023 against the will of the company, which ran to arms flaunting copyright protection to protect its LLM.