Generative AI (GenAI) and large language models (LLMs) present unprecedented ethical challenges, necessitating a reexamination of established AI ethics principles. This paper synthesizes emerging GenAI ethics literature, identifying key themes, knowledge gaps, and future directions. Employing a scoping review methodology, the authors examine how established AI ethics principles are adapted for GenAI and propose new principles to address its unique capabilities and risks. The findings reveal two main themes: the enduring relevance of principles like fairness, privacy, and non-maleficence, and the emergence of new principles tailored to GenAI, including respect for intellectual property, truthfulness, and human-centric design. Despite these contributions, significant gaps remain, such as balancing competing ethical principles, addressing risks like misinformation and malicious use, and operationalizing ethics for diverse stakeholder contexts. The study underscores the need for dynamic ethical frameworks and further research to guide the responsible development and deployment of GenAI technologies.
Read the full paper here.
Citation: Laine J., Minkkinen M., Mäntymäki M. (2025) Understanding the Ethics of Generative AI: Established and New Ethical Principles