Enrique Dans's Medium article argues that generative AI, such as ChatGPT, synthesizes information rather than copies it, and therefore does not infringe on copyright laws. He asserts that the sheer volume of data used to train these models makes it unlikely that any individual source would be plagiarized.

Dans also suggests that copyright law needs to be updated to accommodate generative AI, which is a natural extension of previous artistic creations, and that it is unreasonable to try and restrict access to publicly available internet content for the purposes of algorithm training.

He concludes by expressing hope that a recent court ruling dismissing a copyright infringement lawsuit against OpenAI will set a precedent for future cases.

This article is also available in English on my Medium page, «At last, it seems the law recognizes that AI synthesizes, not copies«