Despite the increasingly common usage of AI to improve work performance, little is known about the perception of individuals who receive assistance from AI, as opposed to a human. Using physician-patient communication as the context, the present research found that AI-assisted physicians were perceived as less competent than human-assisted physicians. In addition, participants reported a lower likelihood of scheduling appointments with AI-assisted (versus human-assisted) physicians and evaluated them less positively. However, this negative effect on perceived competence was moderated by individuals’ beliefs in occupational specialization. Participants with higher beliefs in occupational specialization, where different professions have expertise in specific areas, rated human-assisted physicians as more competent than AI-assisted physicians. Conversely, this difference was absent among participants with lower levels of belief in occupational specialization. These findings highlight the importance for healthcare providers and policymakers to consider the potential impact of disclosing AI assistance. It also has broader implications for domains where human-AI collaborations are used to enhance job performance and perceptions of the individuals involved in such collaborations are important.