Expert: AI training does not make online content free to use

Lawyer Javier López argues that the notion AI may freely train on any online content is unfounded, since rightsholders retain exclusive rights under Article 17.1 of Spain's IP law, the EU InfoSoc Directive and Section 106 of the US Copyright Act. Disputes have reached US courts, including The New York Times v. OpenAI and Microsoft; Munich Regional Court ruled in GEMA v. OpenAI that text and data mining cannot justify storing and reproducing protected works.

• Wikimedia files brief in AI copyright cases2026-10-30
• Why Is Sam Altman Still a Free Man?2026-09-29
• Docs show Microsoft, OpenAI called scraping theft2026-09-17
• Tech firms battle surge of AI-discovered bugs2026-10-28
• News Organizations Challenge OpenAI and Microsoft Fair Use Claims2026-10-23