On 3 November 2026, an adviser to the Court of Justice of the European Union (CJEU) is scheduled to deliver an opinion in the Google Gemini copyright dispute. The opinion is expected to interpret the application of the EU Copyright Directive to the use of copyrighted material for training large language models. News publishers and AI developers across the EU will be directly impacted, with the opinion potentially limiting the scope of permissible data harvesting for AI training. It will also set parameters for future cross‑border enforcement of copyright in AI.

Full News Breakdown

The dispute arose after a Hungarian news publisher alleged that Google’s Gemini chatbot reproduced its articles without permission. The core disagreement centered on whether the text‑and‑data‑mining exception in the copyright framework covers the systematic scraping of news content for AI model training. The matter was referred to the EU’s top court, and an adviser will now issue an opinion on the legal questions involved.

• Case Name: Google LLC v. Hungarian news publisher (Gemini AI)

• Court: Court of Justice of the European Union

• Date: Opinion to be delivered 3 November 2026

• EU Instruments / UK Legislation Cited: Directive 2001/29/EC (the Copyright Directive)

• Key Provisions: Article 5(3) – text‑and‑data‑mining exception; Article 3(1) – exclusive rights of reproduction

• Primary Legal Issue: Scope of the TDM exception for training large language models without the rights‑holder’s consent

• Applicant/Plaintiff Arguments: Unauthorised copying of protected articles constitutes infringement of the reproduction right

• Respondent/Defendant Arguments: Use of the works falls within the TDM exception for scientific research and AI development

• Court's Reasoning: To be provided in the forthcoming opinion

• Holding: Not yet rendered; opinion will guide the final judgment

• Operative Order: Adviser to deliver opinion on 3 November 2026

• Practical Outcome: Pending clarification of lawful data acquisition for AI training across the EU

How Does This Affect You?

Before the opinion, practitioners faced uncertainty about whether large‑scale text scraping for AI training could rely on the TDM exception. The CJEU adviser will now delineate the boundaries of that exception with respect to copyrighted news content. This clarification will make the legality of data‑harvesting practices more certain for AI developers while increasing compliance risk for publishers asserting infringement. The following sections translate those shifts for distinct professional audiences.

For Lawyers & Advocates

• Review and, where necessary, amend existing data‑acquisition contracts to include explicit licences for any content that falls outside the clarified TDM scope, because reliance on the exception alone may no longer be defensible.

• Conduct a gap analysis of clients’ AI training pipelines to identify any scraped news feeds that lack a licence, focusing on the specific “systematic and large‑scale” criterion highlighted by the opinion.

• Draft cease‑and‑desist letters that reference the forthcoming interpretation of Article 5(3) as a basis for demanding retroactive licences from AI developers who have used protected articles.

• Prepare evidentiary bundles for future infringement actions that demonstrate the distinction between “research‑only” use and commercial model training, leveraging the opinion as a benchmark for admissible arguments.

• Advise clients on the limited defensive value of the TDM exception in cross‑border disputes, noting that the opinion may narrow the exception to activities that are “non‑commercial” and “strictly scientific,” thereby increasing exposure to enforcement actions.

For Law Students

The case illustrates how EU courts balance the public‑interest exception for text‑and‑data‑mining against exclusive reproduction rights.
The central doctrinal tension lies between Article 5(3) of the Copyright Directive and the exclusive rights protected under Article 3(1).
The decision is particularly relevant for the study of:

• Copyright law and the TDM exception

• EU internal market and digital single market policy

• Competition law implications of data monopolies

• Comparative intellectual property regimes

• AI ethics and regulatory compliance

Comparable cases include Infopaq International A/S v. Danske Bank (C‑5/08, 2009) and Public Relations Consultants Association v. Google (C‑210/13, 2015); contrasting them with the Gemini opinion highlights how the court’s approach to “systematic extraction” evolves over time.

For Businesses

• Media companies must reassess licensing strategies for AI‑training datasets, ensuring that any future agreements explicitly cover large‑scale text mining to avoid infringement claims.

• AI developers should implement a compliance checkpoint in their data‑pipeline architecture that flags any scraped content from news outlets lacking a licence, thereby preventing unlawful use before model training.

• Boards of technology firms need to approve a revised data‑governance policy that incorporates the clarified limits of the TDM exception, as failure to do so could trigger collective actions by rights‑holders.

• Procurement teams must update vendor due‑diligence questionnaires to capture whether suppliers’ data‑feeds respect the newly defined boundaries of lawful text mining.

Key Takeaways

• The CJEU adviser’s opinion will concretise that the TDM exception does not automatically cover commercial AI training that uses copyrighted news articles without a licence.

• Lawyers must now embed licence verification steps into AI‑related transactions and adjust infringement risk assessments accordingly.

• Regulators and courts will be able to enforce copyright claims against AI developers more readily, while the scope for invoking the TDM defence will be narrowed.

• Watch for the European Commission’s forthcoming “AI Training Data Regulation” expected in early 2027, which will likely codify the opinion’s guidance into statutory rules.

• In‑house counsel should audit all AI‑training data sources before the end of Q1 2027 to ensure compliance with the clarified exception and avoid potential litigation.

Source: Google’s Gemini copyright case back on EU top court calendar