A/78/310 circumstances or different results should obtain different types of explanations”. 50 Additionally, it has been noted that: the explainability of artificial intelligence is an aspiration that is understandable from an ethical and even legal point of view, but it has profound technical difficulties that are worth knowing and, probably, a large part of the solution will also be technical, to the extent that it is possible to redesign algorithms or identify new ones that satisfy ethical and regulatory aspirations. 51 UNESCO, for its part, has indicated that: Explainability refers to making intelligible and providing insight into the outcome of [artificial intelligence] systems. The explainability of [artificial intelligence] systems also refers to the understandability of the input, output, and functioning of each algorithmic building block and how it contributes to the outcome of the systems. 52 58. In order to determine the scope of the principle of explainability, it is necessary to bear in mind its objective and, on that basis, to establish what is needed to achieve it. In line with the above, it has been pointed out that: if the principle of explainability is intended for any hum an being to know why a decision is made based on the processing of his or her data with [artificial intelligence] tools, then the explanation should at least be clear, simple, complete, truthful and easily understood by the person requesting the explanation. It is not enough to report on the data used as inputs to generate the decision, rather, the logic or methodology used to reach the decision should be provided. The challenge is not minor, but it is achievable if there is a willingness to easily explain to people why a decision was generated based on the processing of their personal data. 53 59. Below are some examples of local laws in countries that have tacitly or explicitly incorporated the principle of explainability into their legal frameworks. 60. In Colombia, the law prohibits the processing of data that “misleads” 54 and, in the specific case of decisions made with respect to loan applications, requires those who reject such applications to inform the person concerned in writing, if so required, of “the objective reasons for the rejection”. 55 61. In Ecuador, article 20 of the Data Protection Organic Act establishes that data owners, faced with a decision based solely or partially on assessments resulting from automated processes, including profiling, that produce legal effects on them or that violate their fundamental rights and freedoms, may demand a reasoned explanation of the decision, obtain the assessment criteria on the automated program, submit observations, request information on the types of data used and the source from which __________________ 50 51 52 53 54 55 18/20 Gavilán, Ignacio, “Cuatro principios para una buena explicabilidad de los algoritmos” (2022). Available at: https://ignaciogavilan.com/cuatro-principios-para-una-buena-explicabilidad-de-losalgoritmos/. Ibid. See https://unesdoc.unesco.org/ark:/48223/pf0000381137 p. 23. Nelson Remolina Angarita, “Del principio de explicabilidad en la inteligencia artificial (notas preliminares)”, in Protección de datos personales: doctrina y jurisprudencia , Pablo Palazzi, ed., vol. III (Centre for Technology and Society, University of San Andrés, Buenos Aires, 2023). Statutory Act No. 1581 of 2012, which establishes general provisions for the protection of personal data, art. 4 d). Statutory Act No. 2157 of 2021, which amends and supplements Statutory Act No. 1266 of 2008 and establishes general provisions on habeas data in relation to financial, credit, commercial, service and third-country information and other provisions, art. 5, para. 1. 23-15851