Argentina: Argentina: Disposición 2/2023 — Recomendaciones para una Inteligencia Artificial Fiable

1. ARTÍCULO 3° - Establécese que la presente entrará en vigencia desde su publicación en el BOLETÍN OFICIAL.
2. ARTÍCULO 4° - Comuníquese, publíquese, dése a la DIRECCIÓN NACIONAL DEL REGISTRO OFICIAL y archívese.
3. ANEXO I
4. 1. Consideraciones preliminares
5. 1.1 Objetivo
6. 1.2 Alcance
7. 2. Marco conceptual
8. 3.2.1 Conformar un equipo humano diverso y multidisciplinario
9. 3.3 Aspectos éticos a considerar dentro del ciclo de la IA
10. 3.3.1. Punto de partida común para el equipo diverso y multidisciplinario
11. 3.3.2 Consideraciones éticas respecto al diseño de los datos
12. 3.3.3 Consideraciones éticas respecto del diseño de los modelos
13. 3.3.5 Ética de datos
14. 4. Glosario
15. ANEXO II: DI-2023-61679227-APN-SSTI#JGM

The outline comes from headings detected in the text, so it might miss some or include extras.

1. ARTICLE 3 - It is established that this measure shall enter into force upon its publication in the OFFICIAL G…
2. ARTICLE 4 - Let it be communicated, published, sent to the NATIONAL DIRECTORATE OF THE OFFICIAL REGISTRY, and…
3. ANNEX I
4. 1. Preliminary considerations
5. 1.1 Objective
6. 1.2 Scope
7. 2. Conceptual framework
8. 3.1 How is it advisable to conceive of artificial intelligence?
9. 3.2 What is recommended to do before starting the AI lifecycle?
10. 3.2.1 Establish a diverse and multidisciplinary human team
11. 3.2.2 What is the existing level of awareness in the organization?
12. 3.2.3 Is the use of artificial intelligence indispensable for the problem to be solved?
13. 3.2.4 What is the scope of the basic models for AI adoption?
14. 3.2.5 What is the intended use of the AI, and how will human control be carried out?
15. 3.2.6 What is a premortem analysis?
16. 3.2.7 What now...?
17. 3.3 Ethical aspects to consider within the AI cycle
18. Stage No. 1: Design and data modeling
19. 3.3.1. Common starting point for the diverse and multidisciplinary team
20. 3.3.2 Ethical considerations regarding data design
21. 3.3.3 Ethical considerations regarding model design
22. Stage No. 2: Verification/Validation
23. 3.3.4 How are the specific ethical skills necessary for the AI project validated?
24. 3.3.5 Data ethics
25. 3.3.6 How are the aspects relating to the ethical design of AI models validated?
26. 3.3.7 How are the verifications/validations recorded?
27. Stage No. 3: Implement
28. 3.3.8 How to establish an adequate degree of information security?
29. 3.3.9 What aspects must be taken into account when establishing traceability?
30. 3.3.10 What aspects must be taken into account so that the systems are auditable?
31. 3.4 What ethical aspects should be considered outside the AI lifecycle?
32. 4. Glossary
33. ANNEX II: DI-2023-61679227-APN-SSTI#JGM

The outline comes from headings detected in the text, so it might miss some or include extras.