<p>The increasing popularity of chatbots has led to multiple ventures to create artificially intelligent companions to serve as their users’ friends. In this article, I argue that examining the role of friendship in Aristotle’s moral philosophy and Hegel’s notion of recognition allows us to see why artificial intelligence chatbots cannot replace human friendship. Aristotle grounds friendship in virtue rather than mere utility or pleasure and argues that this form of friendship is essential to a fulfilling life. To understand why this kind of friendship is necessary, I explore Hegel’s concept of recognition, stressing the need for mutual recognition with another self, and how Aristotle’s notion of virtuous friendship aligns with the concept. Despite the attempts by chatbot providers like Replika.ai and Friend.com to alleviate common barriers to the equal consideration of chatbot–human friendship, using Aristotle and Hegel’s philosophy, we can see the incommensurability between chatbot and human friendships. This allows us to understand why a chatbot which is always in a subservient position to its user cannot replace human friendships. The article also considers what it would take for a human–AI friendship to be worthwhile for a person, finding chatbots designed to be people’s friends cannot fulfill this function in satisfactory manner. The article concludes with a reflection on how chatbot friendships could exacerbate, rather than alleviate, loneliness.</p>