Client Alert- Mexico Advances Landmark Reforms to Federal Labor Law and Federal Copyright Law Addressing Artificial Intelligence and Performers’ Rights

Critical Compliance Implications for AI Companies, Technology Platforms, Content Producers, Advertisers and All Businesses Operating in Mexico

 

Executive Summary

On February 23, 2026, Mexico’s President Claudia Sheinbaum submitted to Congress a Presidential Initiative proposing sweeping amendments to both the Federal Labor Law (Ley Federal del Trabajo, or LFT) and the Federal Copyright Law (Ley Federal del Derecho de Autor, or LFDA). The bill was referred to the Committee on Culture and Cinematography of the Chamber of Deputies, which has now issued a formal dictamen (committee report) approving the initiative with modifications, clearing the path for a plenary vote.

The proposed decree is the most significant legislative effort in Mexico to date to regulate the intersection of artificial intelligence with labor rights, copyright, image and voice rights, and the creative industries. Although framed primarily as a measure to protect dubbing actors and performing artists, the reform’s reach extends far beyond the entertainment sector. Its provisions would directly impact AI developers, technology platforms, advertisers, audiovisual producers, broadcasters, and virtually any business using AI tools that interact with voice, image, or artistic performances.

The initiative arrives with a legitimate promise but also a material risk: it promises to shield performing artists from unauthorized AI exploitation, but its broad drafting may push the legal system toward a logic of preventive suspicion where nearly all technology-driven reuse of creative content appears presumptively unlawful.

 

Background: The Initiative and Its Legislative Path

The initiative was sent to the Chamber of Deputies by the Executive Branch on February 23, 2026, and formally referred to the Committee on Culture and Cinematography on February 24, 2026. The Committee consolidated the presidential initiative with eight related legislative proposals filed between May 2025 and March 2026 by legislators.

The Committee received opinions from the Labor and Social Welfare Committee (favorable), the Center for Legal Studies and Parliamentary Research or CEDIP (generally favorable but with substantive technical observations on definitional precision and operational feasibility), and several individual legislators who offered a mix of support and concern. The committee report was approved in a positive sense with modifications, and the text is now ready for plenary debate.

Given the breadth of cross-party support, the reform has strong prospects for approval.

This reform does not exist in isolation. Mexico’s regulatory posture on AI and intellectual property has been evolving rapidly. In 2024–2025, both the Federal Administrative Court (TFJA) and the Supreme Court (SCJN) established that works generated autonomously by AI systems cannot be registered for copyright protection, since authorship requires human creativity under Article 12 of the LFDA. The current reform addresses the other side of the coin: not whether AI can be an author, but whether AI can be used to exploit human performers’ rights without consent.

 

Key Takeaway 1: New Prohibition on AI-Based Reproduction of Performers

The Core Provision. The centerpiece of the labor law reforms is a new article, which provides that—absent express, free, and informed consent coupled with remuneration between the parties—the total or partial reproduction of performing artists through computational technological tools or any other related means, known or yet to be developed, is prohibited when the reproduction serves purposes beyond the originally contracted performances (functions, seasons, or time units).

Three elements are structurally important: (a) the consent standard is elevated to express, free, and informed—ruling out implied, tacit, or boilerplate consent; (b) remuneration is not optional but a condition of validity of the authorization; and (c) the “known or yet to be developed” clause is designed to future-proof the prohibition against emerging AI technologies including voice cloning, deepfakes, and generative models trained on performer data.

Sanctions. Violations carry fines of 1,000 to 5,000 UMAs (Unidades de Medida y Actualización). At current 2026 UMA values (approximately MXN 113.14 per day), this translates to fines of approximately MXN 113,140 to MXN 565,700 (roughly USD 5,600 to USD 28,000) per infraction. Repeat offenses double the fine, up to twice the maximum. These are without prejudice to additional sanctions under the LFDA.

The Risk. As analysts have observed, the text is considerably broad for a digital and technological environment where the boundary between legitimate use and improper exploitation is not clear-cut. When the law prohibits reproduction “by technological means” without carefully distinguishing between abusive cloning, legitimate restoration, authorized dubbing, parody, satire, preservation, or accessibility, the result may be a brake on innovation rather than a targeted remedy.

 

Key Takeaway 2: Voice Recognized as a Component of Image Rights

The reform replaces the concept of “portrait” (retrato) with the broader concept of “image, including voice” (imagen, incluida la voz), extending protection to the characters a person interprets. Use of a person’s image or voice in any format, model, or AI system—or the publication of outputs generated by such systems—requires express authorization, which can be revoked. When a person receives specific remuneration for use of their image, consent is presumed only for the expressly agreed purposes and modalities; any different use requires new authorization and new remuneration. These rights subsist for 50 years after death.

The purpose is to prevent a person’s identity from becoming raw material available to any digital clone. However, the drafting is broad enough to generate more uncertainty than certainty—particularly for businesses that use AI tools interacting with voice or image data in non-entertainment contexts such as customer service, training, or accessibility.

 

Key Takeaway 3: New Exclusive Right to Control AI Transformation of Performances

The reform adds a seventh exclusive right for performing artists: the right to authorize or prohibit the transformation, modification, or imitation of their performances by means of an AI model or system. This creates a new category of exclusive right specifically targeting AI-driven manipulation of artistic performances, a right that did not previously exist in Mexican law or, notably, in international treaties to which Mexico is a party.

Separately, the default license implied by audiovisual production contracts now explicitly excludes the right to generate images or sounds based on the performer’s interpretation through AI, unless expressly agreed. Existing contracts lacking such express authorization will not support AI-based exploitation.

 

Key Takeaway 4: AI Programs as Protected Computer Programs

AI programs are expressly included within the category of “computer programs” protected under copyright law. However, the reform adds a carve-out: AI programs that violate third-party rights protected by the LFDA are excluded from copyright protection. This creates a dual framework—protecting legitimate AI development while denying the copyright shield to AI systems used as instruments of infringement.

The uncertainty identified by observers is significant. Extending to AI programs the same legal regime as literary works does not mean attributing authorial quality identical to a classical literary work. If the legislature conflates tool with work, it creates a grey zone where software is protected but also surrounded by legal and creative expectations that do not correspond to its nature. Industry groups such as ALAI (the Latin American Internet Association) have warned that this formula could disincentivize open-source models and push developers toward excessive protective measures from the design stage.

 

Key Takeaway 5: Dubbing Actors and Voice-Over Artists Expressly Recognized

The definition of “performing artist” is expanded to expressly include “announcers, commercial announcers, and dubbing actors” (locutor, locutor comercial, actor de doblaje). The Committee’s report characterizes this as the structurally most important modification of the entire initiative, closing a gap exploited by producers and platforms to deny these professionals recognition as holders of neighboring rights and access to collective management mechanisms and royalties.

 

Key Takeaway 6: Advertising Contract Reforms and Tighter Exclusivity Rules

The reforms modernize advertising contracts to cover digital media, platforms, geolocation, territory, and temporality. The period after which continued use of advertising materials requires rights holder re-authorization is reduced from three years to one year, with mandatory inflation-adjusted compensation. Exclusivity or non-compete clauses cannot extend beyond the contract’s term. Any modification to contract conditions must result from a free, informed, remunerated, and transparent agreement.

Additionally, a new pro-performer interpretive rule is established: in cases of doubt regarding contractual clauses, the interpretation most favorable to the performing artist prevails.

 

Key Takeaway 7: Alternative Dispute Resolution and Institutional Modernization

The reform replaces the single “avenencia” (conciliation) procedure before INDAUTOR with four voluntary mechanisms: avenencia, mediation, conciliation, and arbitration, available in-person, online, or in mixed format. This aligns the LFDA with the General Law on Alternative Dispute Resolution Mechanisms (LGMASC) of January 2024. The reform also allows INDAUTOR to conduct personal notifications by email or other electronic means, and creates a new category of rights reservations for artistic and cultural events (festivals, fairs, competitions).

A concern raised by observers is that in collective rights management, associations tend to occupy the center of negotiation and collection, while individual artists often arrive with little margin and a pre-determined contract. A good law must protect the performer, not merely strengthen the intermediaries that administer their repertoire.

 

What This Means for Your Business

 

For AI Developers and Technology Companies

Companies developing or deploying AI models that process, replicate, clone, or generate voice, image, or artistic performances will need to implement robust consent and licensing frameworks before training models on performer data or generating synthetic content based on specific individuals. The “known or yet to be developed” language means this obligation cannot be designed around through technological evolution. AI programs used to violate third-party rights under the LFDA lose their own copyright protection, a significant deterrent. The absence of a legal definition of “model or system of artificial intelligence” (acknowledged by the Committee itself) creates additional compliance uncertainty that companies should address by adopting broad interpretations for internal purposes.

 

For Audiovisual Producers and Streaming Platforms

The explicit exclusion of AI-generated content from default audiovisual production contracts means that any use of AI to generate or modify performer likenesses or voices must be separately negotiated. Existing contracts without express AI authorization will not support such exploitation. Platforms should audit their existing talent agreements immediately.

 

For Advertisers, Agencies, and Brands

The reforms require significantly more granular contractual specifications—platforms, geolocation, territory, temporality, and format. The reduction of the re-authorization period from three to one year, combined with mandatory inflation adjustments, increases ongoing costs for long-running campaigns. Exclusivity clauses are strictly limited to the contract term. AI-generated modifications to advertising content featuring performers require separate, specific consent and remuneration.

 

For All Businesses Operating in Mexico

Any entity using AI-powered tools that interact with voice or image data—from customer service systems using cloned or synthetic voices, to marketing departments using AI-generated imagery based on real persons, to HR departments deploying AI interview tools—should review their practices against the new consent and remuneration requirements. The reform’s broad language (“any technological computational tool or any other related means”) is not limited to entertainment applications. The ambiguity of terms like “publish” and the burden of permanent surveillance this could impose on platforms and developers deserve careful legislative attention.

 

Recommended Next Steps

1. Audit existing contracts. Review all agreements with performing artists, voice actors, dubbing talent, commercial announcers, and musicians to identify gaps in AI-specific authorizations. Contracts that do not expressly authorize AI-related uses will not support such exploitation under the reformed framework.

 

2. Implement express, informed consent protocols. Develop consent forms and contractual clauses that specifically address AI uses, training, cloning, transformation, imitation, and synthetic generation—with clear descriptions of each modality and separate remuneration provisions. The consent standard is express, free, and informed; boilerplate language will not suffice.

 

3. Update advertising agreements. Revise advertising contracts to include platform, geolocation, territory, temporality, and format specifications. Calendar re-authorization and inflation-adjustment obligations at the one-year mark. Remove or restructure exclusivity clauses that extend beyond contract terms.

 

4. Assess AI model training data. If your AI models have been trained on voice recordings, performances, or image data of identifiable individuals in Mexico, evaluate whether you have obtained the required level of consent. The reform’s implications for existing AI models trained on Mexican performers’ data remain an area of legal uncertainty that should be monitored closely.

 

5. Register performer contracts. The addition of dubbing, voice-over, and commercial announcement contracts to the list of registrable instruments creates an opportunity for evidentiary protection. Registration with INDAUTOR provides date certainty and third-party enforceability.

 

6. Monitor the legislative timeline. The bill has cleared committee and awaits a plenary vote in the Chamber of Deputies, followed by Senate consideration. The decree would enter into force the day after publication in the Official Gazette (DOF), with a 60-day window for regulatory amendments to the LFDA’s implementing regulations. The Committee has expressly recommended that a legal definition of “model or system of artificial intelligence” be addressed in subsequent legislation.

 

7. Distinguish protection from over-regulation. The line between protecting performers from abusive cloning and penalizing legitimate innovation, including research, parody, preservation, accessibility, and authorized dubbing, is where the real compliance challenge lies. Companies should distinguish between copying a performer to replace them and using a tool to produce a new work with significant human intervention, and structure their AI governance accordingly.

 

Conclusion

The proposed reforms to the LFT and LFDA represent a watershed moment in Mexico’s regulation of artificial intelligence and creative rights. The initiative rightly identifies a real and urgent problem: the unauthorized exploitation of performers’ voices, images, and interpretations through AI systems that can replicate human artistry at scale. Mexico’s explicit recognition of dubbing actors, voice-over artists, and locutors as protected performers, and the creation of a new exclusive right over AI transformation of performances, are genuine advances that set Mexico apart from most jurisdictions internationally.

At the same time, the reform’s broad drafting creates real compliance challenges for companies across sectors. The prohibition on technological reproduction “known or yet to be developed,” the expanded image rights covering AI outputs, and the undefined term “model or system of artificial intelligence” will require careful interpretation by INDAUTOR, by courts, and by businesses designing their AI governance frameworks. The task ahead is to protect creators without petrifying technology—to prevent real abuse, not every possible future use.

Companies operating in Mexico should begin preparing now, before the plenary vote, to ensure that their contracts, consent protocols, and AI deployment practices are aligned with what will be a significantly more demanding regulatory environment.

For additional information, please contact:

Sergio Legorreta at [email protected] with any questions or more specific situations.

 

FisherBroyles is an international law firm practicing in a number of jurisdictions both in the United States and overseas through affiliated legal entities and branch offices of those entities.  Legal services in Mexico are provided through Bravo Gutierrez & Münch, S.C., a member of FisherBroyles (the “Contracting Member”), with offices located in Mexico City, at Parque Lincoln, 5th Floor, Aristoteles 77, Polanco, Mexico City, Ciudad de Mexico 11560 and in Monterrey, at Blvd. Antonio L. Rodriguez 3000-5to piso Interior, 501 Torre Albia, Col. Santa Maria 64650 Monterrey, N.L.

The FisherBroyles Members engage in coordinated international legal practice and may share certain support services but are separate legal entities, each of which is solely responsible for its own work and is not responsible for the work of any other FisherBroyles Member.  Each FisherBroyles Member is subject to the laws and regulations of the particular jurisdiction or jurisdictions in which it operates. Full details of the legal and regulatory status of each FisherBroyles Member are available on the FisherBroyles website.

The use of the name FisherBroyles is for description purposes only and does not imply that the Member Firms are in a partnership or are part of an LLP.  The use of the word “partner” on any Member Firm’s website or in any other Member Firm materials refers to a partner or member of a FisherBroyles Member or an employee or consultant with equivalent standing and qualifications.  You agree that your relationship is with the Contracting Member and not with another FisherBroyles Member unless otherwise confirmed in writing to you. You also agree that your relationship is not with any individual who is a member, employee, or consultant (including anyone we call a partner) of the Contracting Firm Member, who will therefore assume, to the extent permitted by law, no personal liability to you.  Absent the explicit agreement and consent of both entities involved, no FisherBroyles Member is responsible for the acts or omissions of, nor has any authority to obligate or otherwise bind, any other FisherBroyles Member.