To claim copyright protection, must the author of a creation be human? Confirming case law, a federal trial court ruled that works created by artificial intelligence (AI) are not eligible for copyright protection when the algorithm acted without human intervention.
The case pitted computer scientist Stephen Thaler against the United States Copyright Office (USCO) – the federal agency responsible for registering and administering copyrights in the United States. In 2018, USCO refused to register an image generated by an AI named Dabus, designed by Stephen Thaler. The latter presented the work, entitled A Recent Entrance to Paradise (A recent entry into paradise), like a “ autonomous creation of Dabus » : no human contribution had weighed on the final result.
In a March 2020 letter, the USCO explained that an AI-generated work does not have the necessary human authorship to claim copyright protection. The court supported this position, saying that the Copyright Act 1976 requires a “ creator with the ability to work intellectually, creatively or artistically ». The judge recalled that American law has never “protected from works generated by new technologies operating without the supervision of a human hand”.
It should be noted, however, that this judgment only applies to artwork created by an AI without human intervention. When AI is used as a tool to assist in the creative process, it is possible under certain conditions to obtain copyright, according to a (non-binding) notice published in March 2023 by the USCO. The Copyright Office clarifies that works on which AI would have been used “sufficiently creatively so that the work produced constitutes an original work of the author” may be covered on a case-by-case basis.
If the American legal arsenal does not protect the product of generative AIs, it is possible that it will evolve in the years to come to protect human creations against the misuse of AIs. For several months, screenwriters and actors in American cinema have been demanding guarantees to protect their works and their professions against companies that develop generative AI models. Last July, the US Federal Government signed an agreement with the seven largest AI companies – Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI – to address the many risks posed by this new technology. An agreement considered insufficient by many creators.
In France, the Intellectual Property Code provides that only a natural person can have authorship or a legal person for a collective work. The European Parliament has started work on a first regulation on AI. The European text seeks to establish a uniform legal framework allowing innovation via AI, while guaranteeing security and user rights. Adoption is scheduled for late 2023, early 2024, while full application of the text is scheduled for 2026.