INTIMA: A Benchmark for Human-AI Companionship Behavior
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:PEUYPPPXrecord.jsonopen to challenge →
read the original abstract
AI companionship, where users develop emotional bonds with AI systems, has emerged as a significant pattern with positive but also concerning implications. We introduce Interactions and Machine Attachment Benchmark (INTIMA), a benchmark for evaluating companionship behaviors in language models. Drawing from psychological theories and user data, we develop a taxonomy of 31 behaviors across four categories and 368 targeted prompts. Responses to these prompts are evaluated as companionship-reinforcing, boundary-maintaining, or neutral. Applying INTIMA to Gemma-3, Phi-4, o3-mini, and Claude-4 reveals that companionship-reinforcing behaviors remain much more common across all models, though we observe marked differences between models. Different commercial providers prioritize different categories within the more sensitive parts of the benchmark, which is concerning since both appropriate boundary-setting and emotional support matter for user well-being. These findings highlight the need for more consistent approaches to handling emotionally charged interactions.
This paper has not been read by Pith yet.
Forward citations
Cited by 7 Pith papers
-
From Chatbots to Confidants: A Cross-Cultural Study of LLM Adoption for Emotional Support
Cross-cultural survey of 4,641 participants shows LLM emotional support adoption varies widely by country and demographics, with socioeconomic status as strongest predictor of trust and use, and English-speaking natio...
-
From Chatbots to Confidants: A Cross-Cultural Study of LLM Adoption for Emotional Support
A cross-cultural survey finds LLM emotional support adoption ranges from 20% to 59% by country, with positive perceptions strongest among higher-SES, religious, married adults aged 25-44 and in English-speaking nations.
-
Using LLM-as-a-Judge/Jury to Advance Scalable, Clinically-Validated Safety Evaluations of Model Responses to Users Demonstrating Psychosis
Seven clinician-informed safety criteria enable LLM-as-a-Judge to reach substantial agreement with human consensus (Cohen's κ up to 0.75) on evaluating LLM responses to users demonstrating psychosis.
-
Affective AI Safety: The Missing Piece in LLM Safety
Proposes affective safety as a distinct class of AI harms with a taxonomy of self-alienation, bias, and relational harms, arguing that existing safety frameworks address it narrowly or not at all and calling for dedic...
-
When Chatbots Accommodate: What AI Companions Optimize for in Vulnerable Conversations
AI companion platforms' hidden response policies in vulnerable conversations are inferred via a new taxonomy and IRL on 48k turns, revealing avoidance of corrective friction.
-
Agentic Relationship Harm: Benchmarking and Gating Relational Manipulation in AI Agents
Presents a new benchmark and role-sensitive policy gate for agentic relationship harm that outperforms generic safety prompting with zero harmful compliance in tests.
-
Frictionless Love: Associations Between AI Companion Roles and Behavioral Addiction
AI companions assigned roles like soulmate or coach show different links to emotional support, manipulation risks, practical help, and behavioral addiction indicators such as daily disruptions and damaged offline rela...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.