Mexico protects databases through a robust combination of copyright, trade secret doctrine, contracts, and privacy regulation, without a formal sui generis database law. The Federal Copyright Law (LFDA, by its Spanish acronym) protects original compilations under Article 107 and grants database owners reproduction, adaptation, and distribution rights under Article 110. Distinctively, Article 108 protects non-original databases against unfair competition — a quasi-sui generis remedy covering investment-heavy compilations such as customer lists, price indices, or technical manuals that fail the originality threshold, though it remains underutilized in litigation.
Trade secrets are governed by Article 163 of the Federal Law for the Protection of Industrial Property, covering confidential information of industrial or commercial application that confers competitive advantage, with administrative, civil, and criminal liability — including imprisonment of two to six years — for violations, subject to exceptions for independent discovery and reverse engineering. Contract law under the Civil and Commercial Codes is a critical tool for assigning ownership, limiting access, and defining permitted uses, including the scope of data use in AI model training.
Personal data protection is a constitutional right under Article 16, implemented through the LFPDPPP, which grants individuals ARCO rights (access, rectification, cancellation, and opposition) and restricts international transfers, with the Secretaría Anticorrupción y Buen Gobierno as supervisory authority (successor to INAI, dissolved 2025). Mexico has no AI-specific legislation, leaving ownership of AI-generated data unresolved, though the National Artificial Intelligence Agenda for 2024–2030 aims to address regulatory gaps and promote responsible AI development. The Federal Criminal Code additionally criminalizes unauthorized access to and theft of databases.
Key takeaways
- LFDA Article 108 grants the maker of a non-original database a five-year exclusive-use right — a rare neighbouring right in the Americas.
- Trade secret violations can trigger criminal penalties of two to six years' imprisonment, plus injunctive relief through the Mexican Institute of Industrial Property.
- Machine-generated outputs are not copyrightable; human authorship in curation or labeling must be shown for AI-related datasets.
- Data contracts should address AI training scope, retention, and derivative rights, and must comply with the LFPDPPP for personal data.
- No comprehensive AI law exists; the ANIA National AI Agenda 2024–2030 remains a non-binding proposal, while the 2026 LFDA and LFPPI reforms introduced Mexico's first AI-specific IP provisions.
Primary legal instruments
AI-assisted summary of the report, editorially reviewed before publication · AI transparency
Protection mechanisms
Report-derived analysis per mechanism. The report text is fixed; committee members with contributor access can revise the rating (with an audit trail) and add notes below each block.
Recent Developments
Live updates contributed by committee members and the AI research bot — separate from the fixed report content above.
Mexico amends Copyright and Labor Laws to protect performers' voice and image from AI cloning
A decree published in the Diario Oficial de la Federación on May 14, 2026 amends the Federal Copyright Law and Federal Labor Law to require express consent before the image or voice of performers, or their characters, may be used or published, expressly covering outputs generated by artificial intelligence systems. Employment contracts for performing artists must now specify conditions and remuneration for AI-based uses of their image or voice, with exceptions for parody and satire and substantial fines for unauthorized digital cloning. The reform also confirms that computer programs, including AI software, are protected on the same terms as literary works.