MEXICO Law and Practice Contributed by: Eduardo Kleinberg, Adolfo Athié, Claudio Ulloa and Juan Carlos Hernández, Basham, Ringe y Correa S.C.
6.6 Reverse Engineering Reverse engineering is generally lawful in Mexico, par- ticularly when conducted on products that have been lawfully acquired and are available on the market. In such circumstances, it is typically considered a legiti- mate method of obtaining information and does not, by itself, constitute infringement of intellectual prop- erty or trade secret rights. Contractual restrictions may nevertheless prohibit reverse engineering, usually through licence agree- ments, terms of use, or confidentiality clauses. These provisions are generally enforceable under Mexican contract law, provided they are validly agreed and do not conflict with mandatory legal rules. Their practical effectiveness depends on the existence and enforce- ability of a binding contractual relationship between the parties. 7. Data, AI and Emerging IP Issues 7.1 Data Rights and Database Protection A database that is original due to the selection or dis- position of its contents can be protected just as a copyrighted work. Databases that are not original may still be subject to a five-year right of exclusive use. Mexico also protects information that is confidential and provides a competitive advantage as industrial secrets. This information could feasibly include data- sets, provided the owner implements appropriate measures to prevent disclosure and the information provides a competitive advantage. In addition, local criminal codes often sanction the disclosure of infor- mation that is obtained in the course of professional activities, which may also include data sets. Protection through non-disclosure or similar agree- ments is also possible. The approach to scraping and data extraction is the same as for any other kind of work or confidential information. There are no specific defences or excep- tions applicable to it.
7.2 AI-Generated Works and Inventorship/ Authorship In Mexico, an author or inventor is always an indi- vidual, and an AI cannot be considered as an author or inventor. The courts have validated this restriction. In Mexico, it is therefore not legally possible to name AI as an author or inventor. Regarding AI outputs, the Mexican Supreme Court has held that AI creations without a human contri- bution are in the public domain and not protected. The Court considers that the creation of a work is a purely human activity and that a human contribution is needed for the work to be protected. A business would need to clearly understand the com- ponent of a work created with the assistance of AI that are attributable to human contribution in order for the work to be protected and to enforce its copyright in Mexico. 7.3 Training Data, Model Development and Infringement Risk Copying, communicating or making derivative works of a copyrighted work without authorisation may con- stitute copyright infringement and subject the infringer to severe fines, claims for damages, injunctions, etc. Unlike other jurisdictions, Mexico does not have a general “fair use” defence to copyright infringement. If the output resulting from the use of trained data- bases is similar enough to be considered a derivative work, there may be potential for liability. There are no specific defences or exceptions for data-mining. If the data mined includes personal data such as the names, image and likeness of identifiable individuals, then the use of that data requires the consent of the individual, unless they are taken from a publicly avail- able source or subjected to a dissociation process. Industrial secrets cannot be used or exploited without the consent of the respective owner. Any misappro- priation, undue access or disclosure could be subject to criminal prosecution, even if made with the use of AI tools.
226 CHAMBERS.COM
Powered by FlippingBook