Escrever uma avaliação PREreview

Current applications of generative AI technologies—large language models and diffusion models-based picture generators—have been linked with profound ethical and societal issues. Misinformation, intellectual property hijacking, algorithmic bias, and content integrity problems must be well-regulated. Existing frameworks address these issues in disconnected and missing a clear actionable framework. This study conceives a broad-ranging ethical AI-content creation regulation regime, founded on three pillars: top-level ethical guidelines, process working norms derived from guidelines, and technical facilitation mechanisms. The regime emphasizes transparency, accountability, fairness, and human governance with implementation avenues through data regulation, explainability, content provenance, and user interaction. Guided by positive research methodology and synthesis of literature-based research, the idealized regime redresses mistakes in existing regulatory and technology practice. Policy recommendations are presented for policymakers, developers, and users as inputs to be used by them for accountable AI output generation, deployment, and use. Incorporating ethical theory and functional ICT mechanisms, the project is evolving on reliable, fair, and human-centered AI systems.