The French regulatory authority has fined Google €250 million for violating EU copyright laws in AI training. The corporation announced this in a blog post.
The regulator expressed concerns about the AI service Gemini, claiming that the model was trained on content from local publishers and news agencies without proper notification and permission.
In the blog, Google acknowledged the accusation but emphasized that it is the “first and only platform with licensing agreements with 280 French press publishers.”
The company described the fine as “disproportionate” to the violations noted by the regulatory authority. It also stated that the regulator did not sufficiently recognize Google’s efforts to respond to warnings.
The fine is related to a copyright dispute in France, initiated due to complaints from several leading news companies in the country, including Agence France Presse.
The disagreements emerged shortly after the corporation, in February, jointly with the French government, announced plans to establish a new artificial intelligence center in the country.
It will be based in Paris and will employ 300 researchers and engineers.
Earlier, Reuters reported that Google signed a $60 million agreement with Reddit, allowing the company to train its AI models on the platform’s content. The agreement could serve as a model for future contracts of a similar nature.
Back in February, The New York Times filed a lawsuit against OpenAI and Microsoft for using articles to train chatbots without its permission. In response, Sam Altman’s company accused NYT of “hacking” ChatGPT to fabricate evidence.