George RR Martin, creator of “Game of Thrones”
A Manhattan judge has agreed to move forward with the lawsuit by writer George RR Martin, creator of “Game of Thrones”, against OpenAI, considering that there is sufficient evidence that ChatGPT was trained with his work without authorization.
A court decision involving the American writer George R.R. Martin and the use of Generative Artificial Intelligence marks a turning point in the debate on copyright and technology.
According to , a Manhattan judge authorized the action brought by the writer to proceed to trial, after considering that there is sufficient evidence that Martin’s novels may have been used to train ChatGPT without your permission.
The case gained prominence when Martin’s lawyers asked OpenAI’s LLM to creation of a fictional sequel to “A Clash of Kings”having received in response a work entitled “Dance of Shadows”.
The text incorporated new characters and narrative elements from the universe of A Game of Thrones, which, according to the authors of the action, demonstrated that the system would have learned directly from the original books by Martin.
Legal action and arguments of the authors
The complaint was filed through the Authors Guilda union that brings together several writers who denounce a massive appropriation of his works to train models without receiving compensation.
Among the signatories are recognized prestigious authorswho argue that this practice constitutes a misuse of protected works for copyright.
Months before the action, several creators had already warned companies technologies for “injustice inherent in the exploitation of our works as part of its AI systems, without our consent, credit or compensation.”
This notice anticipated the growing concern in the literary sector given the ability of models to reproduce styles and weaves.
The conflict is part of a global context in which different countries maintain divergent interpretations of “legitimate use”.
A recent example was the extrajudicial settlement of 1.5 billion euros achieved by Anthropic with a group of authors, avoiding a definitive court ruling on the unauthorized use of their texts in training AI models.
The controversy contrasts with the UK Supreme Court rulingwhich concluded that the Stability AI did not violate copyright by training its models with images from Getty Images. This divergence of criteria uncertainty increases about how legal limits on technological development should be defined.
The debate is not centered concept of legitimate usethe main argument of technology companies, which argue that training their models constitutes a transformative process, notes .
On the other hand, writers maintain that this activity replaces the original work and violates fundamental rights of the creator.
The final decision in the George RR Martin case could influence future regulation of the sector and determine how AI models will be built in the coming years.
