As AI companion apps gain widespread popularity, research questions whether user attachments to chatbots fulfill developmental psychologist Mary Ainsworth’s six conditions for a true attachment bond, scientificamerican.com reported.
Defining Attachment Through Ainsworth’s Six Conditions
Attachment is the term psychologists use for the most significant relationships that hold people’s lives together. According to the pioneering American-Canadian developmental psychologist Mary Ainsworth, an attachment relationship must meet six precise conditions. The relationship must persist over time, having been formed with a particular subject who cannot be replaced. The relationship must also be emotionally significant, involve the desire to remain close, and cause distress at involuntary separation. Finally, when threatened, the person must seek security and comfort in their relationship with that subject.
The first five conditions describe any lasting affectionate bond, such as a close friendship. Only the sixth turns it into an attachment bond specifically. Simply liking someone, feeling close to them, or spending hours together is not enough to create an attachment relationship. Interestingly, nothing in these six conditions specifies a human on the other side. Long before the emergence of conversational AI chatbots, deities, pets, and objects ranging from children’s blankets and stuffed animals to cherished household items, phones, and computers have all been found to supply security and comfort when someone feels distressed.
What Current Research Actually Measures
Today, AI companion apps are commonly sold on the promise of lasting and close relationships. Many users describe feeling real grief when an app is updated or withdrawn. But do such experiences meet the narrow definition of attachment?
So far, studies exploring human-AI interactions have documented how often people use chatbots, their emotional responses toward them, and their behaviors when deprived of access. It has not tested these relationships comprehensively against Ainsworth’s six conditions. Researchers have created, validated, and implemented questionnaires designed to measure attachment to AI, yet most of these tools seemingly presume the bond is indeed an attachment from the outset. Public discourse and academic literature alike have widely adopted terminology centered around attachment, and even hyper-attachment.
Data from a recently released poll spanning 7,027 participants across Germany, China, South Africa, and the U.S. indicated that a minimum of one-third of individuals utilizing AI chatbots exhibit patterns suggesting an attachment bond. However, its three questions asked whether people consider the chatbot a friend, are polite to it, and miss it after a few days. At most, those questions touched on emotional significance and mild separation anxiety. None asked whether people turn to chatbots for security and comfort when truly distressed.
Why Distinguishing AI Comfort From True Attachment Matters
Whether relationships between humans and AI chatbots count as attachment matters because the word travels. It shapes how clinicians and parents read behavior, how regulators frame consequences, and how companies describe their products.
A large portion of the existing data and public conversations regarding human-AI chatbot interactions relies on metrics tracking usage frequency and subjective emotional reactions. Yet, such measures cannot separate attachment from other processes. People report feeling less lonely after chatting with an AI companion, which resembles seeking comfort. However, as some researchers have pointed out, available studies did not assess whether emotions are actually regulated through the chatbot, thereby conflating distress relief with attachment.
Separation anxiety identical to that experienced when losing a person can occur when users part with an AI companion. But that condition applies to all affectionate bonds and does not, on its own, identify attachment. Calling these relationships attachments therefore goes further than the evidence currently allows.
Contrasting Findings and Open Questions in Human-AI Studies
The evidence against attachment is not conclusive either. Although this could suggest that relationships with AI are not attachments, the same mismatch has been found for pets, and pets can serve as attachment figures. Thus, this mismatch alone does not rule out attachment.
The practical question remains whether AI use helps or harms.
