Previous research suggests that participants may be susceptible to confirmation bias after making decisions in moral dilemmas. We manipulated the type of moral dilemmas (personal or impersonal) and the framing of the question prompting participants to respond (emphasizing saving five people or sacrificing one person). The actors in the dilemmas were represented by a series of silhouettes. Eye tracking data revealed that both manipulations had an effect on participants' gaze. Further analysis of utilitarian choices has shown that there were no framing effects of the prompting question when the dilemmas were impersonal. The data suggests that participants' subsequent gaze patterns are sensitive to both how the situation is described and the framing of their hypothetical actions. Taken together, our results provide some support to the claim that confirmation bias may arise after making moral decisions.