Establishing ethical design principles for ‘Humanized’ generative AI bots in workplace coaching conversations

The purpose of this conceptual paper is to explore the design principles which can be used to inform the design and deployment of AI coaching agents, an area of growing interest and activity. The paper draws on coaching psychology, AI ethics and empirical studies of AI coaching to offer seven design principles:(i) transparency and informed continuous consent, (ii) autonomy and user empowerment, (iii) privacy and data protection, (iv) fairness, (v) do no harm and beneficence, (vi) accountability and (vii) alignment. Further it argues that generative AI tools should be viewed not as agency free but as more akin to a domesticated pet and that liability for unintended negative events, such as harm to a person, should be shared between the designers, the IP holders and the AI agent. The paper aims to inform this emerging agenda for designers, organisational buyers and coaching psychologists who are considering or deploying AI workplace coaching agents.

‍