On the Troubled Relation Between AI Ethics and AI Safety

Kapitel i bok, 2026

The terms “AI ethics” and “AI safety” are fuzzy around the edges and overlap, but roughly speaking the former addresses the more incremental and contextual effects of present-day AI, while the latter puts more emphasis on transformative consequences of future AI, including existential risk. There is a perceived conflict between the two fields, centered around the idea of a competition for a joint but fixed pool of attention and resources. This chapter argues that the perception is based mostly on misunderstandings, and that if there is to be a conflict at all in this sphere, it makes much more sense if AI ethicists and AI safetyists join forces against the so-called AI accelerationists.