In an escalating battle between Google and French regulators over copyright protections for news content, the tech giant has been slapped with a staggering €250 million ($270 million) fine by the country’s competition authority. The Autorité de la Concurrence announced the penalty, accusing Google of disregarding commitments made to news publishers regarding the use of their copyrighted material.
The fine stems from Google’s controversial use of news publishers’ content to train its generative AI model Bard, now known as Gemini, without properly notifying the copyright holders. This decision is particularly notable as it shines a spotlight on the emerging challenges surrounding the use of copyrighted data to develop powerful AI systems.
A Prolonged Tussle over Fair Compensation The dispute between Google and French authorities dates back to 2019 when the European Union extended copyright protections to news snippets and headlines. News aggregators like Google News were then required to negotiate fair compensation with publishers for reusing their content.
Initially, Google sought to bypass the law by shutting down Google News in France, a move that prompted the competition authority to intervene, deeming it an abuse of Google’s dominant market position. This forced Google to the negotiating table with local publishers.
However, in 2021, Google faced a €500 million ($592 million) fine for breaching commitments made during those negotiations. The tech giant appealed the sanction but later withdrew its appeal and offered a series of new pledges to resolve the dispute.
Failure to Notify and Discriminatory Practices In its latest decision, the Autorité de la Concurrence found that Google had used content from news publishers and agencies to train its AI chatbot Bard/Gemini without properly notifying the copyright holders or the regulator itself. This violated Google’s commitment to provide transparency to publishers regarding the use of their content.
The authority also criticized Google for failing to provide a technical solution until September 2023 that would allow publishers to opt-out of having their content used for AI training without affecting its display on other Google services.
Beyond the AI-related issues, the regulator called out several other shortcomings in Google’s dealings with publishers, including:
- Opaque methodology for calculating remuneration to publishers
- Failure to meet non-discrimination criteria, ensuring equal treatment
- Imposing a arbitrary minimum threshold for remuneration, discriminating against smaller publishers
- Incorrect calculations regarding “indirect income” from content reuse
- Failure to update remuneration contracts as per previous commitments
A Costly Saga with No End in Sight While Google has agreed not to appeal the latest fine in exchange for a faster resolution process, the company’s managing director for news partnerships, Sulina Connal, struck a defiant tone, stating that “the fine is not proportionate to the issues raised.”
However, with generative AI rapidly evolving and the competitive scramble to launch powerful models, Google’s calculus on the content reuse issue appears to have shifted. As the tech industry grapples with the ethical and legal implications of training AI on copyrighted data, this saga is likely far from over.
The substantial €250 million penalty serves as a stern warning from French authorities, signaling their unwavering commitment to protecting the rights of news publishers and ensuring fair compensation in the digital age. As the lines between AI development and copyright infringement blur, regulators worldwide will be closely watching this high-stakes battle unfold.