Navigating the New Era: Regulation of Artificial Intelligence in Mexico.

In Mexico, legislation on affairs of international interest, such as the use of artificial intelligence, often sparks discussion and interest among both lawmakers and the private sector players. On February 28, Senator Ricardo Monreal caught attention by introducing his federal law initiative to the Senate of the Republic aimed at regulating artificial intelligence.

The proposed law seeks to regulate the development, marketing, and use of artificial intelligence (AI) systems in Mexico. Its goal is to ensure respect for human rights, protect intellectual property rights, and foster the national development of AI systems. Furthermore, the proposal aims to establish an ethical and responsible framework for AI use, promoting transparency and preventing potential negative impacts on society.

This will be monitored by the Federal Telecommunications Institute, which presently operates with four out of the seven commissioners. Additionally, a consultative entity is being established, designated as the National Commission for Artificial Intelligence, which is expected to comprise five eminent scientists in the field. The concern arises not from a lack of experts in the country, but from the lack of clarity in the selection process.

The initiative, with its 24 articles and six transitional provisions, contains ample material to raise concerns. Its primary objective, with the best of intentions, is to safeguard the integrity of Mexicans. The arguments presented and the cases cited in its statement of reasons, when compared to international legislation, support this objective. However, this regulation requires a thorough evaluation, as it only superficially addresses the issues it intends to legislate, such as certain copyright matters or the use of AI for audiovisual modifications. An example is the case of a financial worker deceived by scammers using deepfake technology, resulting in a loss of 25 million dollars to his company¹. Furthermore, the proposal acknowledges the potential for AI to be used for malicious purposes, such as manipulation and violation of fundamental rights. AI is classified according to the risk it may pose, including arguments that could affect the development of AI in Mexico. An example is the classification and prohibition of so-called unacceptable risk treats. Although these protective measures appear clear, the definitions are ambiguous. If the state keeps the decision to unilaterally reclassify AI, there is a risk that it may exceed its protective or monopolistic intent and carry out these reclassifications at will.

A significant point is the proposed obligations for developers, such as the need for systems to undergo evaluation and human control as determined by the authority, and the requirement to seek prior authorization before each product launch². These obligations could slow down development and participation in a market where time is critical, resulting in additional costs that usually are absorbed by the final consumer. Moreover, it is proposed that the developer disclose elements that could be considered industrial secrets, that can be considered competitive advantage. If this bill is implemented, it would require revisiting the protection of methodologies through patents, as is done in other countries.

In conclusion, the artificial intelligence regulation bill presented by Senator Monreal represents a significant step towards establishing a legal framework for this field in Mexico. The initiative acknowledges the need to legislate on this issue. However, like the proposal of May 23, 2023, for the ethical regulation of artificial intelligence³, it faces modernization challenges and may result in hurdles for AI developers. This underscores the importance of not only creating specific legislation for AI but also modernizing and expanding the general legislative framework to properly integrate this new bill.

1. Finance worker pays out $25 million after video call with deepfake ‘chief financial officer’. Source: https://edition.cnn.com/2024/02/04/asia/deepfake-cfo-scam-hong-kong-intl-hnk/index.html

2. A regulation like HIPAA can incur fines of over $250,000 for a single violation in the international case. In 2020, French data protection authorities fined Google and Amazon $120 million and €42 million respectively for failing to request permissions from customers to drop non-essential cookies. Source: https://selleo.com/blog/the-importance-of-compliance-and-security-in-software-development

3. Source: http://sil.gobernacion.gob.mx/Archivos/Documentos/2023/04/asun_4543395_20230413_1680209417.pdf