Who Embraces the Machine? How Familiarity, Mastery and Context Shape Attitudes towards AI Involvement

The rapid rise of generative Artificial Intelligence (genAI) has intensified calls to flag its involvement in everyday content, but it remains unclear who welcomes such assistance and when disclosure sustains favorable evaluations. We collected data from 6,362 participants, aged 18–65, across six socio-economically and technologically diverse countries. Attitudes varied by task type: genAI was favored for logistical tasks, whereas humans were preferred for interpersonal advice. Higher AI-use frequency, self-reported expertise, and sense of mastery predicted warmer baseline attitudes when disclosure occurred upfront; these same traits magnified backlash when disclosure was delayed, consistent with identity-threat and expectancy-violation accounts. The findings identify a hidden risk: post-hoc labels can alienate the very audience most inclined to champion genAI. Early, task-tailored disclosure, paired with user-controlled editing options, therefore offers a path toward calibrated trust and responsible deployment, underscoring the need to tailor transparency not only to what AI does, but also to who is using it and when they learn the truth.