Dutch journalists' unions and authors launch Opstand.ai campaign and prepare collective lawsuit against Meta over AI training

What happened: On 29 September 2026, the Dutch journalists' union NVJ, the collective management organisation Lira Foundation, and the Dutch authors' union Auteursbond held a demonstration outside Meta's Amsterdam headquarters as the launch of the Opstand.ai campaign, and announced they are preparing a collective lawsuit against Meta. The European Federation of Journalists publicly backed the action, with EFJ leadership holding a parallel event in Brussels.

The groups allege Meta used the work of freelance journalists, writers and translators to train its AI models without consent and without pay, including works allegedly obtained from illegal shadow libraries. Their stated goal is a collective agreement with AI companies that ensures fair pay, gives creators control over whether their texts are used, and requires transparency about which works are used for training.

Why it matters to creators: this is one of the first organised cross-sector collective actions in Europe against an AI company over training data, and it frames the fight around consent, compensation, and transparency rather than fair-use litigation alone. The shadow-library angle, if litigated, could test the EU Text and Data Mining exception's lawful-access requirement for generative AI training.

What is not yet known: no lawsuit has been filed yet — the action was announced as in preparation. Whether the case will proceed, what claims it will bring, and how Meta will respond remain open.