Mapping the Ethics of Generative AI: A Comprehensive Scoping Review

Abstract

The advent of generative artificial intelligence and the widespread adoption of it in society engendered intensive debates about its ethical implications and risks. These risks often differ from those associated with traditional discriminative machine learning. To synthesize the recent discourse and map its normative concepts, we conducted a scoping review on the ethics of generative artificial intelligence, including especially large language models and text-to-image models. Our analysis provides a taxonomy of 378 normative issues in 19 topic areas and ranks them according to their prevalence in the literature. The study offers a comprehensive overview for scholars, practitioners, or policymakers, condensing the ethical debates surrounding fairness, safety, harmful content, hallucinations, privacy, interaction risks, security, alignment, societal impacts, and others. We discuss the results, evaluate imbalances in the literature, and explore unsubstantiated risk scenarios.

References in corpus (19)

• Artificial Intelligence: the global landscape of ethics guidelines
• Principles alone cannot guarantee ethical AI
• Art and the science of generative AI: A deeper dive
• A Survey on ChatGPT: AI-Generated Contents, Challenges, and Solutions
• Normalizing Flows for Probabilistic Modeling and Inference
• Managing extreme AI risks amid rapid progress
• ChatGPT: Vision and Challenges
• A Taxonomy of Prompt Modifiers for Text-To-Image Generation
• Conformity Assessments and Post-market Monitoring: A Guide to the Role of Auditing in the Proposed European AI Regulation
• Smiling Women Pitching Down: Auditing Representational and Presentational Gender Biases in Image Generative AI
• AI virtues -- The missing link in putting AI ethics into practice
• Deception Abilities Emerged in Large Language Models
• Speciesist bias in AI -- How AI applications perpetuate discrimination and unfair outcomes against animals
• Decoding the Threat Landscape : ChatGPT, FraudGPT, and WormGPT in Social Engineering Attacks
• Deceptive AI Ecosystems: The Case of ChatGPT
• The Ethical Implications of Generative Audio Models: A Systematic Literature Review
• Forbidden knowledge in machine learning -- Reflections on the limits of research and publication
• Visual Adversarial Examples Jailbreak Aligned Large Language Models
• Negative Human Rights as a Basis for Long-term AI Safety and Regulation

Cited by in corpus (4)

• The AI risk repository: A meta-review, database, and taxonomy of risks from artificial intelligence
• Generative AI as a Playful yet Offensive Tourist: Exploring Tensions Between Playful Features and Citizen Concerns in Designing Urban Play
• Large Reasoning Models Are Autonomous Jailbreak Agents
• When Image Generation Goes Wrong: A Safety Analysis of Stable Diffusion Models