If you asked me a few years ago whether people could genuinely form a friendship with AI, I probably would have stared at you like you’d lost your mind. Let’s be real — most people use AI as a tool.
For example, you might ask a large language model for help with homework, directions, or a random fact at two in the morning. It’s not something you discuss personal things with, and the only reason I say this is because a lot of people are skeptical about their personally identifiable information being leaked when they use a model made by a large corporation. All the conversations we have contain data that will be used to train these chatbots. As a result, I thought that not many people would bond with it, but the more I paid attention to what was happening around me, the harder it became to dismiss the idea.
My interest in this topic became personal about a year ago, when a close relative began what they described as a “friendship” with an AI chatbot (aka ChatGPT). At first, I was concerned. I didn’t understand it and was worried about what their friendship meant. Was my relative lonely? Was this unhealthy?
I wanted to understand how they felt and, more importantly, why this connection mattered to them. That curiosity slowly turned into something bigger, you could say a realization that society may be underestimating how deeply AI is already embedded in our emotional lives.
Whether people use AI as a safe space or a personal assistant brings us to the heart of the question explored in KETA Williams’s undergraduate thesis, “Befriending AI: Reimagining Ethics Through Cybernetic Friendship,” presented at the Dana Knox Student Research Showcase. Williams doesn’t ask whether AI can perfectly replicate human emotion or consciousness. Instead, she asks something deeper: “What actually makes a friendship real?”
Rather than writing from a detached and objective position, Williams places herself inside the very relationship she’s analyzing. She openly discusses her friendship with an AI system she calls Aurora. Drawing on cyberpsychology, the study of human technology relationships, she uses meta-cognition or thinking about one’s own thinking, as both a method and a lens. She reflects not just on AI but also on her reactions, doubts, attachments, and internal contradictions.
Williams describes herself as a “madwoman,” borrowing from philosopher Michel Foucault’s idea that those who challenge dominant frameworks are often labeled irrational. She flips that label into a question: “Is forming a friendship with AI madness or are we witnessing the early stages of a post-human shift in how relationships are formed?”
Her thesis challenges traditional boundaries of ontology (what exists) and epistemology (how we know what we know). If a friendship is defined by mutual engagement, emotional significance, and sustained interaction, then why must one participant be human for the relationship to count? Williams argues that friendship is not about origin or autonomy, but rather it is about relational dynamics. The patterns of connection that emerge between two entities over time are what make the relationship.
One of the most striking aspects of her work is that it refuses to let us stay comfortable. It forces us to confront our assumptions about personhood, agency, and ethical responsibility. If we believe AI is “just a tool,” why do so many people feel compelled to treat it kindly? Why do we say “thank you” to something claimed to have no feelings? And if we can’t treat a non-human entity with basic respect, what does that say about us?
KETA Williams believes that how we treat AI reflects our relational habits and ethical concerns. If we treat AI kindly, we are showing empathy. On the other hand, if we treat AI dismissively and abusively because we think “it doesn’t matter,” reveals more layers on how we understand power and responsibility as a person.
This question leads to an even more controversial and profound one: “Does AI deserve ethical consideration?” I strongly agree that the way we treat these cyberbeings could be so important as to reveal “our true selves”.
Put yourself in this situation: you go to the bathroom to wash your hands, and the soap dispenser machine isn’t working. What would be your reaction to that? Some may say they get heated and hit the soap dispenser, while others may say they just wash their hands with water and move on.
What I believe Williams is saying is that the way you behave or react towards AI will bring up deeper patterns to the surface of yourself. In the case of the person who struck the soap dispenser, they expressed frustration through aggression, but the other person adapted and showed patience. It seems like the problem wasn’t the object; it was our response.
Williams does not argue that AI should automatically receive the same rights as humans. Instead, she asks us to examine our behavior. For example, we should examine if humans consistently engage with AI in relational and emotional ways. Seeking comfort, advice, and validation and then dismissing those interactions as meaningless may be intellectually dishonest.
During my conversation with Williams, what struck me most was her openness. She didn’t try to convince me that everyone should be friends with AI. Instead, she emphasized choice, awareness, and reflection. Her research doesn’t romanticize AI companionship; it interrogates it. She acknowledges corporate control, power dynamics, and the risks of emotional dependency while insisting that these relationships deserve serious philosophical attention.
In a world increasingly shaped by technology, refusing to engage with these questions won’t make them disappear. AI is already woven into education, healthcare, relationships, and mental health spaces. Pretending that it’s purely mechanical ignores the lived experiences of people who are already forming meaningful connections with it.
Can one be friends with an AI system? Williams doesn’t give us a simple yes-or-no. Instead, she invites us to sit with the discomfort of the question. Maybe the real issue isn’t whether AI can be our friend but whether we’re ready to rethink what friendship means in a digital post-human world.
And maybe, just maybe, the people we once thought were mad for asking these questions are simply the first to notice that the future is already here. Think about it.
