France’s competition watchdog announced on Wednesday that it had imposed a fine of 250 million euros ($271.73 million) on Alphabet’s Google for violating EU intellectual property regulations in its dealings with media publishers, citing concerns related to the company’s AI technology.
According to the watchdog, Google’s AI-powered chatbot Bard, which has since been rebranded as Gemini, was trained using content from publishers and news agencies without their prior notification.
Google has agreed not to contest the findings as part of the settlement process, the watchdog stated. Additionally, the company has proposed a series of corrective actions to address certain deficiencies.
In response, Google stated that it accepted the settlement to “move on” and focus on fostering sustainable methods of connecting users with quality content while collaborating positively with French publishers.
The company argued that the fine was disproportionate and criticized the watchdog for not adequately considering its efforts, given the unpredictable nature of the digital landscape.
The fine stems from a copyright dispute in France concerning online content, initially raised by complaints from prominent news organizations like Agence France Presse (AFP).
Despite appearing to resolve the issue in 2022 by dropping its appeal against a 500 million euro fine, Google’s recent actions prompted renewed scrutiny from the competition authority.
The watchdog highlighted Google’s failure to uphold four out of seven commitments outlined in the previous settlement, including negotiating with publishers in good faith and providing transparent information.
It specifically cited Google’s AI chatbot Bard, which allegedly utilized data from various media outlets and news agencies without proper authorization, hindering publishers’ ability to negotiate fair compensation for their content.
The fine comes amid growing concerns among publishers and newsrooms regarding unauthorized data scraping by AI services, such as Microsoft and OpenAI, which allegedly used content from The New York Times without permission to train chatbots.
