TRACEABILITY AND CERTIFICATION OF AI PROMPTS: A Normative Proposal for the Ethical Governance of Automated Content

The increasing use of generative artificial intelligence has intensified debates around ethical traceability, digital authorship, and governance of algorithmically produced content. This article introduces a novel framework for certifying AI prompts — the fundamental input commands shaping generative outputs — by proposing a normative-technical model inspired by metadata protocols, DOI registration systems, and ISO certification standards. The model integrates legal frameworks such as Brazil’s LGPD and anticipates the establishment of an accredited Prompt Certification Body (PCB), comparable to national standards authorities like ABNT or INPI. Methodologically, the study employs benchmarking, expert consultation, and a simulated case involving legal AI-generated content, demonstrating practical feasibility and regulatory potential. The certification mechanism aims to reassign accountability, prevent misinformation, and enhance algorithmic transparency in fields such as law, education, journalism, and public policy. Ultimately, the proposal advances a new layer of AI governance, bridging creative freedom with normative precision, and fostering a trust-based digital ecosystem grounded in ethical standards and democratic oversight.