Man vill inte att sina hjältar ska vara fusk : En studie i vilka komponenter som kan ligga till grund för en etisk certifiering för AI i film- och spelbranschen

Sammanfattning

This study examines the question of which components an ethical certification of generative AI for the film and game industries should contain. Previous research shows a lack of certification models for specific industries and points to several advantages as well as ethical dilemmas associated with AI technology in both sectors. The study’s research question is answered through semi-structured interviews with industry actors from both industries and is analyzed with the support of the seven key requirements Human agency and oversight, Technical robustness and safety, Privacy and data governance, Transparency, Diversity, non-discrimination and fairness, Societal and environmental wellbeing, and Accountability. The results include components such as that human decision-making needs to be a requirement, that transparency needs to permeate the entire process, and that the responsibility should be assigned to a politically independent state actor. Furthermore, it is proposed that the certification should be based on transparency rather than detailed requirements, with a focus on making AI usage visible to consumers.