Unpacking the Risks of Confiding in AI: A Philosophical and Communication Ethics Perspective

The development of Artificial Intelligence (AI) has transformed the way humans acquire information, work, and communicate. Today, many people use AI as a place to ask questions, seek solutions, and even pour out feelings that are difficult to convey to others. This phenomenon shows that AI is no longer seen merely as a technological tool, but is beginning to act as a conversational companion. Behind this convenience, questions arise that need to be critically examined. Can confiding in AI truly replace human-to-human communication? Does the emotional relationship built with AI retain ethical value? These questions are important because communication is essentially not just an exchange of information, but also a process of building empathy, responsibility, and mutual understanding. As AI develops, issues of transparency, accountability, and privacy protection in digital communication also become significant concerns.

From a philosophical perspective, humans are seen as thinking beings who possess consciousness and are capable of giving meaning to their life experiences. AI can indeed generate quick responses that appear to understand the user’s emotions, but AI does not have consciousness, life experience, or moral responsibility like a human. The responses given are the result of data processing and language patterns, not the result of genuinely felt empathy. This view aligns with the thinking of Martin Buber through his concepts of I–Thou and I–It. According to Buber, an I–Thou relationship is built on the recognition of fellow humans as whole persons, thus giving rise to empathy, trust, and openness. Conversely, an I–It relationship views the other party as an object or tool. In the context of confiding in AI, the relationship formed more closely resembles the I–It concept because AI is merely a technology that processes information without consciousness, life experience, or the ability to feel emotions like a human. Therefore, AI can help provide perspective, but cannot replace authentic interpersonal relationships.

Furthermore, Jürgen Habermas, through his theory of Communicative Action, explains that the goal of communication is to achieve mutual understanding through rational, open dialogue free from domination. Habermas emphasises that understanding is not built by one party alone, but through intersubjective communication that allows each individual to express opinions equally and be open to better arguments. In the context of AI use, this process is not fully realised because AI lacks moral consciousness and the ability to engage in dialogue as an equal subject. Consequently, AI is more appropriately positioned as a tool for reflection rather than a replacement for human-to-human communication. The phenomenon of confiding in AI can also be explained through Social Presence Theory. This theory explains that a person’s perception of the presence of a conversational partner can be influenced by the digital media or interface used. When AI provides fast, personal, and empathetic-sounding responses, users can feel as though they are talking to someone who is truly present. However, this presence is psychological, not a real presence accompanied by consciousness, life experience, and reciprocal relationships as in human communication. As a result, users potentially build emotional closeness that is actually only a perception of the AI system.

From the perspective of communication ethics, there are several risks when a person becomes too dependent on AI as an outlet for feelings. First, the risk of reduced interpersonal communication. When someone is more comfortable talking to AI than to family or friends, social relationships can become increasingly distant. In fact, human communication contains elements of empathy, trust, and emotional support that cannot be fully replaced by technology. Second, there is a risk to user privacy. Personal information shared with AI services must be carefully guarded, especially sensitive data. Third, AI can produce answers that sound convincing but are not necessarily correct, complete, or appropriate to the user’s context. Therefore, AI should not be used as the primary decision-maker in matters of life, health, law, or social relations. This perspective is reinforced by the concept of Ethical AI, which emphasises four main principles: transparency, fairness, privacy and security, and accountability. Transparency requires users to understand that AI works based on data and algorithms, not because it has consciousness or feelings. The principle of fairness reminds us that AI still has the potential to produce bias because it is influenced by training data. The principle of privacy and security emphasises the importance of maintaining the confidentiality of users’ personal information, while accountability requires AI developers to be responsible for the impacts caused by the systems they build. These four principles show that the use of AI must be carried out responsibly so that the benefits of technology do not sacrifice users’ rights and security. Thus, AI is a highly beneficial innovation for obtaining information, discussing, and conducting self-reflection. However, AI cannot replace human-to-human communication.