Intellectual Property 2026

SPAIN Law and Practice Contributed by: Mario Sol Muntañola, Javier Márquez Martín, Ferran Llaquet Ballarín and Ana Padial, Sol Muntañola Abogados

through specific confidentiality clauses included in employment or commercial contracts. Most of these agreements and/or clauses also include a penalty clause, setting out the damages (usually very high sums) in the event of a breach of confidentiality. 6.5 Licensing and Assignment Know-how can be transferred or licensed, just like trade secrets. There are no specific legal provisions governing the form of such transfers or assignments, but they are usually done in writing to provide a clear and precise record of the agreement’s content, with- out any other special formalities. In other words, no special registration is required, nor is it necessary to In Spain, reverse engineering, for the purpose of observing, studying, dismantling or testing a product, is legal provided that the following conditions are met. • Access to the product has been lawful – that is to say, it has been legally acquired on the market. • There is no contract or confidentiality agreement that expressly prohibits reverse engineering of the said object. Any other form of access to the information, through espionage, deception, theft, or breach of a confidenti- ality or non-disclosure agreement, is unlawful in Spain and will be subject, depending on the severity, either to the penalties provided for under civil law or to the more severe penalties under the Criminal Code. have the agreement notarised. 6.6 Reverse Engineering Therefore, if you wish to keep your know-how prop- erly protected, it is advisable to impose contractual restrictions to prohibit reverse engineering. 7. Data, AI and Emerging IP Issues 7.1 Data Rights and Database Protection Within the EU, databases are afforded various forms of protection under intellectual property law. In addi- tion to protection as a work – which is limited solely to cases where their structure is original – the so-called “sui generis” right also protects the substantial invest- ment made by the manufacturer in obtaining, verify-

ing or presenting the content. The presence of this requirement for substantial investment is assessed qualitatively or quantitatively whether in terms of financial resources, time spent, effort or other similar parameters. On the other hand, as regards the protec- tion of the content (the data), this is achieved primarily by treating it as confidential (as trade secrets) and regulating access to and use of it through contracts. Both forms of protection are useful in dealing with the mass, automated extraction of data (for example, web scraping). In short, the protection of databases requires a hybrid approach: copyright (original struc- ture), sui generis rights (substantial investment) and/ or trade secrets/contractual arrangements (dataset). 7.2 AI-Generated Works and Inventorship/ Authorship Generally speaking, results generated exclusively by artificial intelligence (AI) are not protected by copy- right. Under the vast majority of legal systems (as is the case in the European Union, and in Spain in par- ticular), such rights require human authorship. At pre- sent, there is very little specific legislation on AI, and therefore all issues arising from the creation of content using AI must be addressed and resolved under cur- rent intellectual property law, which only recognises authorship for works that are the product of human creation. Given this difficulty in obtaining copyright over works generated by AI, companies’ strategy must involve reinforcing and protecting the added value of the human factor during the creative process. In this regard, to facilitate such protection, it is essential to maintain rigorous documentation of human involve- ment, establish well-defined internal protocols and policies, and make use of contracts and/or trade secrets as a means of protection. 7.3 Training Data, Model Development and Infringement Risk The use of copyright-protected works, or confidential information protected as trade secrets, for the pur- pose of training an AI carries obvious legal risks. If the rights holders are able to prove unauthorised use of their rights for this purpose (which is not always easy), the party responsible may face legal proceedings that could result in a ban on such use and an obligation to

319 CHAMBERS.COM

Powered by