AI-mediated apology in a multilingual work context: Implications for perceived authenticity and willingness to forgive

Abstract

As Artificial Intelligence-Mediated Communication (AI-MC) technology is increasingly used to facilitate communication worldwide, its implications for interpersonal relationships in multinational working environments have become more significant. In particular, knowing that AI-MC tools are used by the communicator might reduce recipients' perceptions of the authenticity of emotionally charged messages, such as an apology. Across three scenario-based studies rooted in an interpersonal work-related conflict, we examined the effects of the choice to use AI-MC tools to communicate an apology, focusing on people's perceptions of the genuineness of the apology, and their ensuing tendency to forgive the person apologizing. We consistently found that the choice to use AI-MC tools diminished perceptions of the apology's authenticity and the consequent willingness to forgive, and that self-disclosing the use of AI-MC on the part of the communicator did not mitigate this effect. However, making limited use of AI-MC (selecting to use only one of three available tools) had no negative impact on the perceived authenticity of the apology, suggesting that limiting the use of AI-MC signals a diminished distance between the original intention of the person apologizing and the final formulation of the message of apology, leading to perceptions of a more genuine apology.