France was the first of the EU’s 27 countries to adopt a copyright directive that allows publishers and newsrooms to enter into licensing agreements with online platforms to use their content, including for AI training. And it was in accordance with this directive that the French antitrust authority Autorité de la concurrence (Competition Authority) fined Google 250 million euros for violating EU rules on the use of intellectual property.

The French regulator said that the AI chatbot Google Bard (Gemini) was trained on the content of publishers and news agencies without notifying them; the corresponding complaints were filed by several news agencies in the country, including Agence France Presse (AFP), and supported by testimony… the Bard chatbot itself, which “admitted” that it was trained on data from unnamed media and news agencies without informing the copyright holder or regulator. The Autorité de la concurrence also noted that Google violated 4 of the 7 conditions of the 2022 pre-trial settlement, under which it undertook to pay 500 million euros, and eliminate a number of shortcomings – in particular, according to the French regulator, it did not provide transparent information and negotiated in bad faith with publishers.

In a blog post, Google said the fine “is disproportionate” in relation to the issues raised and “does not sufficiently take into account” the company’s efforts to address copyright issues, but agreed to pay it “because it’s time to move on.” Given that the amount of recovery has been halved, the movement further away from Google should accelerate.

Літературний редактор.

Commentary