It is widely believed that moral intuitions are unreliable, as they would be subject to numerous and pervasive cognitive biases. In particular, biases in moral intuitions would pose a significant challenge to moral progress, which demands an alignment with objective standards for harm and help. This chapter provides rigorous support to these claims by reviewing the most significant biases in moral intuitions and by arguing that their presence constrains the future improvement of common-sense morality. To this end, after introducing the topic (Sect. 6.1), I first provide a more precise definition of moral intuition bias (Sect. 6.2), followed by a dyadic typology of such biases (Sect. 6.3). I then compare the prototypical forms of harm and help to which moral intuitions are evolutionarily attuned with those that characterize contemporary and future moral challenges, concluding that spontaneous adaptation of moral intuitions to these challenges is unlikely (Sect. 6.4). This analysis supports the normative claim that responsible agents ought to implement targeted debiasing strategies to extend the reliability of moral intuitions beyond their natural constraints.