<p>AI-based systems not only simplify daily life and work but are increasingly used in areas requiring special help, attention, and support. This is especially true in care, where individuals with specific needs and preferences depend on attentive assistance. In response to these demands, AI-based care robots have already been deployed for some time. However, this raises important questions: What happens, for example, if the robot harms a person? Would one blame the robot? Could one blame the robot in an adequate way? And would the machine therefore be morally responsible? This article discusses a specific form of responsibility regarding these machines – namely role responsibility. Although many authors have addressed different versions of (moral) responsibility in the context of care robots, this perspective has been missing from the literature so far. First, the meaning of role responsibility is presented. This is followed by a clarification of how role responsibility differs from moral responsibility. Third, the assumptions made so far are applied to care robots. Lastly, the arguments are summarized and potential objections addressed. This paper argues that while AI-based care robots cannot bear moral responsibility due to their lack of consciousness, intention, and reason-responsiveness, they can still hold role responsibility based on their social function in caregiving contexts. These robots are appropriate targets for (thin) praise and blame, not because they are moral agents, but because they operate within roles tied to morally relevant duties. Recognizing this form of responsibility helps narrow existing responsibility gaps in care environments without displacing human accountability.</p>