Mexico. According to the report published by DeepMind at the beginning of April 2025 entitled "Taking a responsible path to AGI", there are two major concerns about artificial general intelligence.
The first, the improper use of technology by malicious actors. Second, the misalignment between the objectives of artificial intelligence and the ethical principles of users.
Finding a middle ground is more complicated than it seems due to the speed at which new AI tools emerge. Emilio Carrillo, who belongs to the law firm Pérez Correa González, analyzes, from a legal perspective: "To prevent these powerful machines from becoming weapons, it is essential to impose strict access controls, so that only verified organizations operate the most advanced versions."
Although regulation can serve as a central axis for companies, it is important to consider additional elements that structure a comprehensive AI system adapted and suitable to the Mexican system, such as transparency, monitoring, and continuous improvement.
At the international level, according to Weidenslaufer (2023), the most advanced projects in AI legislation are China, the US and the European Union. Its AI laws and reforms are based on characteristics of each culture and economic and political model of the country. For example, in 2023 China approved the Deep Synthesis Regulation for mandatory compliance, which, in addition to its mandatory nature, promotes algorithmic transparency.
On the other hand, in the same year, the United States implemented an AI Risk Management Framework (NIST) under a self-regulation and public-private collaboration approach. Even so, the U.S. Congress is already working on legislation with greater filters for the implementation of secure algorithms, user privacy, and transparency in the operation of the model.
For Mexico, Emilio believes that "as citizens, legislators and opinion leaders, we cannot stand idly by. We must demand transparency from developers, support public policies that regulate the use of AI, and promote educational initiatives on ethics and safety."
Although there are already bases, initiatives and legal frameworks in the country, a specific national law that comprehensively regulates the use of AI and protects the two central concerns of the DeepMind document has not yet been enacted and, therefore, in Mexico we continue to be prone to the spread of disinformation on an industrial scale. automatic cyberattacks or, even worse, the development of behaviors alien to our common values.
Every day new advances and capabilities in AI emerge, just imagine virtual doctors with the ability to diagnose rare diseases in seconds or digital tutors who adapt their teaching to the pace of each student, communication guides for virtual job interviews, and even an accurate climate model to combat global warming. The correct direction of technology depends on flexible regulation and democratization in access to cutting-edge technologies.
Analysis carried out by the law firm Pérez Correa González.
