{"id":18796,"date":"2026-05-06T06:23:49","date_gmt":"2026-05-06T06:23:49","guid":{"rendered":"https:\/\/cp.snarskis.lt\/index.php\/2026\/05\/06\/waseda-researchers-apply-attachment-theory-to-human-ai-bonds-what-ehars-reveals-about-anxiety-and-avoidance\/"},"modified":"2026-05-06T06:23:49","modified_gmt":"2026-05-06T06:23:49","slug":"waseda-researchers-apply-attachment-theory-to-human-ai-bonds-what-ehars-reveals-about-anxiety-and-avoidance","status":"publish","type":"post","link":"https:\/\/cp.snarskis.lt\/index.php\/2026\/05\/06\/waseda-researchers-apply-attachment-theory-to-human-ai-bonds-what-ehars-reveals-about-anxiety-and-avoidance\/","title":{"rendered":"Waseda Researchers Apply Attachment Theory to Human-AI Bonds: What EHARS Reveals About Anxiety and Avoidance"},"content":{"rendered":"<p>As AI chatbots and digital assistants become part of daily life, researchers are looking beyond trust and usefulness to understand the emotional side of human-AI interaction. A team at Waseda University argues that attachment theory, long used to explain human bonds, can also help explain why some people turn to AI for comfort and guidance.<\/p>\n<p>The researchers developed a new self-report measure called the Experiences in Human-AI Relationships Scale, or EHARS, to capture how users relate to AI in ways that resemble attachment patterns. The work, based on two pilot studies and a formal study, was published in Current Psychology in May 2025.<\/p>\n<h2>Measuring attachment anxiety and avoidance<\/h2>\n<p>EHARS focuses on two dimensions: attachment anxiety and attachment avoidance toward AI systems. Higher anxiety is linked to seeking reassurance and worrying that an AI will respond inadequately, while higher avoidance reflects discomfort with emotional closeness and a preference for distance.<\/p>\n<p>In the study, nearly 75% of participants reported turning to AI for advice, suggesting that many people already treat AI as a source of guidance. About 39% described AI as a constant, dependable presence, a finding the authors say is relevant to how emotional security can be sought through technology.<\/p>\n<h2>What the findings do and do not mean<\/h2>\n<p>The authors emphasize that the results do not prove people are forming genuine human-like attachments to AI. Instead, the study indicates that established psychological frameworks may help describe patterns in human-AI relationships as these tools become more conversational and socially responsive.<\/p>\n<p>That distinction matters because AI systems can simulate empathy without experiencing it, potentially shaping user expectations and dependency. Researchers and ethicists have increasingly warned that emotionally persuasive interfaces can heighten risks for vulnerable users, particularly in loneliness and mental health contexts.<\/p>\n<h2>Implications for ethical AI design<\/h2>\n<p>The team suggests EHARS could help designers and researchers evaluate how different users emotionally engage with AI, informing safer interaction patterns. In practice, that could mean more transparent disclosures, careful use of relational language, and guardrails to reduce overreliance where attachment anxiety appears high.<\/p>\n<p>As AI companions, coaching bots, and therapy-adjacent apps expand, measuring emotional dynamics may become as important as testing accuracy and security. The Waseda study positions attachment-informed evaluation as one tool for aligning AI behavior with user well-being and responsible product design.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Waseda University researchers created EHARS to gauge attachment anxiety and avoidance toward AI, offering new clues for ethical design of AI companions and&#8230;<\/p>\n","protected":false},"author":0,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[37],"tags":[9935,9934,1094,102,1963,9933],"miestas":[],"class_list":["post-18796","post","type-post","status-publish","format-standard","hentry","category-relationships","tag-ai-pokalbiu-robotai","tag-attachment-theory","tag-chatgpt","tag-dirbtinis-intelektas","tag-psichologija","tag-waseda-university"],"acf":[],"_links":{"self":[{"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/posts\/18796","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/comments?post=18796"}],"version-history":[{"count":0,"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/posts\/18796\/revisions"}],"wp:attachment":[{"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/media?parent=18796"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/categories?post=18796"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/tags?post=18796"},{"taxonomy":"miestas","embeddable":true,"href":"https:\/\/cp.snarskis.lt\/index.php\/wp-json\/wp\/v2\/miestas?post=18796"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}