News EU JURI Generative AI Copyright Report 2026

From AI Law Wiki
Revision as of 02:34, 28 April 2026 by AILawWikiAdmin (talk | contribs) (Migration export)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

The European Parliament's Committee on Legal Affairs (JURI) adopted a landmark report on generative artificial intelligence and copyright on February 25, 2026, calling for sweeping reforms to EU copyright law to address the unauthorized use of copyrighted works in AI training.[1] The report, titled "Copyright and generative artificial intelligence - opportunities and challenges" (Procedure 2025/2058(INI)), was drafted by Rapporteur Axel Voss and adopted by the JURI committee in a 17-3 vote.[2] The European Parliament subsequently adopted the report in plenary on March 10, 2026, with a large majority.[3]

Background

The report addresses growing legal uncertainties in EU copyright law, including under the Copyright in the Digital Single Market (CDSM) Directive, as generative AI systems have been trained on copyrighted works without licenses, payment, or transparency. The JURI committee found widespread violations including unauthorized scraping, ignoring opt-out declarations, and use of pirated sources.[1][2]

Key Recommendations

AI Training on Copyrighted Works

The report insists that EU copyright law must apply to all generative AI systems on the EU market, regardless of where training took place, referencing the territoriality principle under AI Act Recital 106. Non-compliant AI systems should be barred from the EU market. The report demands respect for rights-holders' refusals to permit training, calls for new licensing frameworks with fair remuneration and collective management, and recommends assessment of past unauthorized uses.[1][4]

Transparency Obligations

GenAI providers and deployers must disclose lists of copyrighted works used, crawling records, and training data details. Non-compliance would create a rebuttable presumption of infringement, potentially leading to legal liability. The transparency obligations extend to inferencing, retrieval-augmented generation (RAG), and all AI deployers.[1][5]

Opt-Out Mechanisms

The report criticizes current opt-out mechanisms as impractical, non-transparent, and incomplete. It affirms that rights-holders can opt out of AI training and data crawling. The European Commission is urged to explore tools for preventing use by general-purpose AI systems and to improve enforcement of opt-out declarations.[4][6]

Implications

The report represents the European Parliament's most comprehensive policy position on AI copyright to date. Key implications include:

  • Market access: AI systems trained on copyrighted works without authorization could face exclusion from the EU market
  • Rebuttable presumption: Failure to disclose training data creates a legal presumption of infringement, shifting the burden to AI companies
  • CDSM review: The Commission is urged to conduct an independent assessment of the copyright acquis before the CDSM review, plus licensing facilitation measures[1][7]

See Also

References