Material de Abuso Sexual Infantil Generado A Partir De La Inteligencia Artificial: Una Revisión de Literatura – Child Sexual Abuse Material Generated from Artificial Intelligence: A Literature Review

This research addresses the question of legal and psychological consequences of child pornography created by artificial intelligence (AI). The problem lies in the misuse of AI to impersonate child images with sexual connotation, generating harmful effects such as violations of human rights and personal integrity, along with legal and psychological repercussions. Rapid technological advancement has created a legal and ethical vacuum. The general objective is to analyze these consequences, while specific objectives include examining current laws, investigating distribution methods and affected populations, and describing characteristics of consumers and distributors. The methodology used is mixed (exploratory-descriptive and social critical), combining documentary analysis, legal hermeneutics, and metrics analysis. The theoretical framework is based on the concept of "Deep Fake," AI-manipulated files that simulate authenticity and often depict children being abused. Generative AI facilitates explicit content, transgressing rights. Virtual Child Sexual Abuse Material (CSAM) can be generated or manipulated by AI. Among the findings, international efforts exist, but current legislation often presents gaps when defining pornography as "real representations," creating ambiguity for artificial content. Platforms such as Facebook, Instagram, and Telegram are used to share this material. Psychologically, the misuse of child imagery causes severe damage including low self-esteem, anxiety, and PTSD symptoms. Consumers and distributors may use identification symbols on the internet. The conclusions indicate a critical absence of specific legal regulation for AI-generated content. Laws such as the Colombian Penal Code define child pornography as "real representations," without encompassing artificial content. The urgent need to reform the Penal Code to include virtual representations, equate sanctions, and establish responsibilities for platforms and AI developers is postulated. An interdisciplinary approach and adaptive regulations are essential.