Background on Gabriela Jiménez Godoy
Gabriela Jiménez Godoy is the Vicecoordinadora (Vice Coordinator) of Morena, a political party in Mexico, serving in the Chamber of Diputados (House of Representatives). Her recent actions have sparked discussions about the regulation of artificial intelligence (AI) in Mexico.
Meeting with Google México and Synergia Asuntos Públicos
According to various news reports published this week, Jiménez Godoy held a working meeting with representatives from Google México and Synergia Asuntos Públicos to seek agreements on AI regulation in Mexico.
AI Regulation Objectives
Jiménez Godoy emphasized that prohibition is not the solution and highlighted two primary objectives for AI regulation: preventing misuse of technology by criminal networks and fostering a sovereign, ethical, and inclusive technological development model.
Critique of the Proposed Legislation
However, upon reviewing the bill she presented in April of this year, concerns arise regarding typical Morena legislative shortcomings. Critics argue that the bill contains excessive political language, abstract and ambiguous concepts, insufficient technical legislative detail, and simulated government control.
Legislative Flaws
The proposed bill, titled “Federal Law for the Ethical, Sovereign, and Inclusive Development of Artificial Intelligence,” displays several legislative vices. The title itself seems more like political propaganda than a formal legal order, according to critics.
Content Analysis
The bill proposes a notably over-regulated environment for AI system developers, introducing concepts like “algorithmic risk,” “algorithmic impact,” “algorithmic audit,” and “algorithmic bias.” These terms aim to increase regulatory burden and developer responsibilities.
Moreover, developers would be obligated to rectify “algorithmic harm” for users affected by “unjust automated decisions” via AI systems. This provision not only exposes the industry to arbitrary criteria for determining unjust automated decisions but also seemingly transfers complete responsibility for user decisions made through AI system interaction to developers.
Ambiguous Restrictions
The bill prohibits AI that uses subliminal manipulation or extreme psychological persuasion techniques to alter human behavior, a restriction so vague that it could encompass common advertising forms. Ironically, the bill also bans AI systems performing mass biometric surveillance in public spaces without judicial authorization, suggesting that only state-sanctioned biometric data collection is valid.
Penalties and Oversight
The bill allows authorities to deactivate an AI system causing algorithmic harm, though the term remains undefined. Penalties include public reprimands, collective damage reparation, and fines of up to 100,000 Units of Measure (over 11 million pesos currently) for developers violating the law. Authorities could also order controlled technical access to a privately-owned model or system.
These punitive and verification measures appear disproportionate, potentially creating a new form of censorship.
Technological Sovereignty and National Development
The bill includes a chapter titled “Technological Sovereignty and National Development,” stipulating preferential treatment for Mexican institutions and ethical principles-based models in public contracts, acquisitions, and projects involving AI.
This preferential treatment is criticized for being ambiguous, potentially inviting corruption, and possibly violating international treaties signed by Mexico.
Key Questions and Answers
- What is the primary concern regarding the proposed AI regulation? Critics argue that the bill contains excessive political language, insufficient technical detail, and vague restrictions that could stifle innovation and potentially create a new form of censorship.
- What are the main objectives of AI regulation according to Gabriela Jiménez Godoy? The objectives are preventing misuse of technology by criminal networks and fostering a sovereign, ethical, and inclusive technological development model.
- How does the proposed bill impact AI developers? The bill imposes heavy regulatory burdens, obligates developers to rectify algorithmic harm, and exposes them to arbitrary criteria for determining unjust automated decisions.
- What are the concerns regarding the bill’s ambiguous restrictions? The vague language could encompass common advertising forms and potentially infringe on privacy rights through mass biometric surveillance.
- What are the potential consequences of the proposed penalties and oversight measures? These measures could disproportionately stifle innovation and create a new form of censorship.
- How might the preferential treatment for Mexican institutions impact international treaties? Critics argue that this preferential treatment could be ambiguous and potentially violate international treaties signed by Mexico.
