The dizzying advance of the artificial intelligence (AI) poses great opportunities, but also ethical, social and legal challenges. Faced with these risks, the Council of Europe adopted in May 2024 el German Framework on Artificial Intelligence, the Rights Humans, the Democracia and the State Law, becoming the first instrument legally binding to level international to regulate the development and use responsible for AI systems.
This treaty lays the foundation for a governance model international based derechos, What are you looking for prevent abuses and to guarantee that AI serves humans and does not replace or marginalize them.
scope of agreement
El German was adopted by the Council of Europe and is open to countries around the world, not only to the members of the Company. Its objective is:
- Ensure that the design, development and use of AI systems respect the derechos human, fundamental liberties, the democracia and the State right.
- Establish principles and obligations binding for the States part, regardless of the sector in which AI is applied (Justice, health, security, public services, education, etc.).
Key principles of the agreement
Among the fundamental pillars of the German Marco highlights:
1. Legality and respect for the derechos human
AI systems must comply with the current constitutional, conventional and international norms. use cannot justify the restriction undue of derechos such as privacy, footing, the freedom of expression or due proceedings.
2. Transparency and explainability
El treaty forces AI systems to be understandable and auditable, especially in automated decisions affecting people. States must to guarantee mechanisms of surrender of counts.
3. Liability and human supervision
AI uses must always provide for the significant human intervention, and regimes of responsibility clear protection against damage or erroneous decisions caused by algorithms.
4. No discrimination y equity
It is prohibited to use of AI that reproduces or aggravates discrimination based on gender, race, ethnicity, origin, edad, disability or sexual orientation. Mitigating algorithmic biases is also required.
5. Access to remedies
All person affected by an AI-based decision must have effective access to resources legalIncluding review by a authority competent.
Obligations of the States
The states part must:
- Adopt national regulatory frameworks compatible with the treaty.
- Establish impact assessments of derechos human before implementing sensitive AI systems.
- Create or strengthen authorities independent technological oversight.
- Ensure the participación public and consultation democratic in the development of AI policies.
- Empowering digital education and algorithmic literacy.
Application in Latin America and the Caribbean
Although the agreement born in Europe, is from universal application. On opening to third countries creates the possibility for States of Latin America and the CaribbeanIncluding Republic Dominican Republic, adhere to it and harmonize their technological policies with the standards of derechos humans.
En particular, can guide:
- El use ethical of AI in judicial and administrative processes.
- La prevention of the mass state vigilantism.
- The regulation of algorithmic recruitment and AI-assisted education.
- Data protection in contexts with biometric technologies.
Un instrument complementary
El German Framework does not seek to replace treaties derechos humans like the Convention of Americana or Pact international de Rights Civilians and politiciansBut complement them in the face of the new challenges posed by technologyArtificial intelligence, as a cross-cutting phenomenon, requires intersectoral and multilateral approaches.
