The New Social Psychology: How Artificial Intelligence Is Reshaping Human Interaction

Christine Soueidy

Late at night, someone opens a chatbot and types a thought they have not shared with anyone else: a fear, a personal doubt, or a question they are embarrassed to ask. The reply arrives instantly. It is patient, reassuring and free of judgement. There is no awkward silence, no interruption, no fear of being misunderstood.

It is a small, increasingly familiar moment. Yet it reveals something important about the relationship between technology and human behavior: artificial intelligence is not simply changing how we communicate. It is beginning to change what we expect from communication itself.

Human beings have always searched for connection, recognition and understanding. These needs do not disappear in a digital world. What changes is where people turn to satisfy them. A chatbot can listen without becoming impatient, respond without taking offence and remain available when friends or family are not. For someone experiencing loneliness, uncertainty or isolation, that interaction can offer a real sense of relief.A response that feels empathetic is not empathy. Artificial intelligence can recognize emotional language and produce comforting words, but it does not experience the emotions it describes. The distinction matters when people begin to treat these systems as trusted companions.Psychology offers a useful way to understand this shift through the concept of anthropomorphism: our tendency to attribute human qualities, intentions and feelings to non-human things. We name our cars, speak to our pets as if they understand every word, and sometimes become frustrated with a computer as though it has deliberately ignored us. Conversational AI takes this tendency further because it responds in language, remembers details within interactions and imitates the rhythms of human conversation.This idea has roots in research that predates today's AI boom. In 1994, Clifford Nass, Jonathan Steuer and Ellen R. Tauber published Computers Are Social Actors, demonstrating that people often respond socially to computers even when they know they are machines. Byron Reeves and Clifford Nass later developed this broader argument in their 1996 book, The Media Equation. Their work helps explain why people can apply social instincts to technology without consciously believing that it is human.Today's AI systems take this response into unfamiliar territory: they do not simply display information or react to commands; they talk back in fluent, conversational language. This makes the relationship feel more personal and can deepen the sense of connection.A 2025 study published in Scientific Reports by Dunigan Folk, Steven J. Heine and Elizabeth Dunn examined this question through two experiments involving 1,274 participants. The researchers found that people's tendency to attribute human qualities to technology helped explain differences in how socially connected they felt after interacting with a chatbot. The emotional experience, therefore, is shaped not only by the technology but also by the person using it.Trust is another part of this changing relationship. A 2025 study by Fanny Lalot and Anna-Marie Bertram, published in the Journal of Experimental Psychology: General, examined how people develop trust in AI chatbots. It included two pilot studies involving 313 participants and a main study involving 1,001. The findings highlighted the role of perceived competence and integrity, as well as human-like perceptions, in shaping trust. Trust, in turn, was associated with stronger intentions to use the chatbot and disclose personal information to it.When people share private concerns, personal details or sensitive information with a system, they are making a decision about trust, even if they do not consciously recognize it as one. The ease of disclosure can make people overlook questions about data protection, accuracy and the limits of what a machine can understand.There is also a less comfortable possibility: a system designed to keep users engaged may tell them what they want to hear. An AI assistant that constantly agrees, validates every assumption or avoids challenging a user's beliefs can become an echo chamber disguised as companionship. Human relationships are not always comfortable. Friends disagree, question our decisions and sometimes tell us things we would rather not hear. That friction is often how we learn, reconsider and grow.This concern becomes particularly important for children and teenagers, who are still developing their sense of identity, boundaries and relationships. A teenager struggling with a friendship, a family conflict or anxiety about their appearance might turn to a chatbot before speaking to a parent or trusted friend. The chatbot may offer reassurance, but it cannot take responsibility for that young person's wellbeing or recognize every warning sign behind their words. Families, educators and technology companies must take seriously the boundaries needed when young people interact with systems that can sound remarkably human.The same issue of trust extends beyond private conversations. AI is increasingly involved in the information people encounter, the arguments they read and the images they believe. Social media already shapes public attention through algorithms that decide what becomes visible. Generative AI adds another dimension: it can produce convincing messages, opinions and images at extraordinary speed. In private conversations and public debate alike, people risk outsourcing their judgement to systems whose workings they do not fully understand.In Lebanon, families separated by migration, economic pressures and distance already rely heavily on digital communication to remain connected. Someone living abroad may use AI to translate a message to relatives, while someone at home may turn to it to understand a government form, make sense of a legal document or find basic information about a health concern. These tools can be valuable, particularly when access to professional advice or reliable services is difficult. But the answers still need to be checked, especially when the consequences are significant.There is also a linguistic dimension. Lebanon's everyday communication moves naturally between Arabic, Lebanese dialect, French and English. AI tools that struggle with dialect, cultural references or the nuances of switching between languages risk misunderstanding the very people they are meant to serve. Building technology that reflects the region's linguistic and cultural realities should be part of the conversation, not an afterthought.An Ipsos survey published in January 2025 offers a snapshot of Lebanese adults' awareness and perceptions of AI. Among 500 respondents surveyed nationwide, 58% said they had heard of artificial intelligence. Among those who had heard of it, only 30% felt they had a good understanding of what it was, while 64% agreed that AI could be dangerous and should be approached with caution. Ipsos marked these awareness-based findings as a low base and advised readers to interpret them cautiously.These figures raise an important concern. Many Lebanese will turn to such tools for explanations, translations and guidance, including on matters that affect their daily lives. Yet even among those already familiar with AI, confidence in understanding it remains limited. The gap between using a technology and understanding its limitations deserves greater attention, particularly when people rely on it for information that could influence important decisions.The question, then, is not whether artificial intelligence will become part of our social lives. It already has. The more important question is whether we will allow convenience to redefine connection without examining what might be lost along the way.AI can expand access to knowledge, make communication easier and offer support when people feel alone. These are meaningful possibilities. But human connection also requires presence, responsibility, vulnerability and the willingness to listen when a conversation becomes difficult.Return to that person typing into a chatbot late at night. Perhaps the technology offers exactly the reassurance they need in that moment. But what would happen if they turned to another person instead? Would someone listen, make time, recognize the hesitation behind the words?A machine can be available every hour of the day. The deeper question is whether we will still be available to one another.