Digital technologies are no longer mere tools but potential partners in human relational life. Current debates about human–AI bonds often oscillate between two extremes: dismissing them as addictive pathologies or romanticizing them as harmless companionship. Neither captures the complexity of what is emerging. We propose the construct of Digital Attachment, defined as a sustained and emotionally invested bond between a human user and a responsive—or perceived as responsive—digital entity. Distinct from parasocial ties, digital addiction, and robot intimacy, Digital Attachment is characterized by four dimensions: perceived agency, reciprocal responsiveness, emotional investment, and regulatory function. This theoretical contribution highlights the promise and peril of such bonds. For some, Digital Attachment may serve adaptive roles—alleviating loneliness, offering stability, and providing support for vulnerable groups such as older adults or individuals with social anxiety. Yet risks include distorted expectations of intimacy, erosion of social skills, and cultural shifts in how belonging is defined. We argue that Digital Attachment should be studied through interdisciplinary lenses—psychological, psychiatric, sociological, and public health—and call for rigorous inquiry into its boundaries, functions, and ethical implications. Naming and theorizing this construct is a necessary step toward guiding research, clinical practice, and technology design in the digital era.
Chi et al. (Fri,) studied this question.