Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RubRIX: Rubric-Driven Risk Mitigation in Caregiver-AI Interactions

Drishti Goel, Jeongah Lee, Qiuyue Joy Zhong, Violeta J. Rodriguez, Daniel S. Brown, Ravi Karkar, Dong Whi Yoo, Koustuv Saha

TL;DR

RubRIX introduces a theory-driven, clinician-validated rubric to characterize and mitigate interactional risks in LLM-generated caregiving responses. Grounded in the Ethics of Care, it defines five risk dimensions and uses a structured evaluator to quantify and guide refinements across six LLMs and two real-world caregiver datasets. RubRIX-guided refinements achieve substantial risk reductions (roughly 45-98%) after a single iteration, with the largest gains in epistemic and normative risks and some model-dependent variability in attentional and factual risks. The work provides a domain-sensitive evaluation framework and releases benchmark data to support safer, caregiver-centered AI deployment in high-burden contexts.

Abstract

Caregivers seeking AI-mediated support express complex needs -- information-seeking, emotional validation, and distress cues -- that warrant careful evaluation of response safety and appropriateness. Existing AI evaluation frameworks, primarily focused on general risks (toxicity, hallucinations, policy violations, etc), may not adequately capture the nuanced risks of LLM-responses in caregiving-contexts. We introduce RubRIX (Rubric-based Risk Index), a theory-driven, clinician-validated framework for evaluating risks in LLM caregiving responses. Grounded in the Elements of an Ethic of Care, RubRIX operationalizes five empirically-derived risk dimensions: Inattention, Bias & Stigma, Information Inaccuracy, Uncritical Affirmation, and Epistemic Arrogance. We evaluate six state-of-the-art LLMs on over 20,000 caregiver queries from Reddit and ALZConnected. Rubric-guided refinement consistently reduced risk-components by 45-98% after one iteration across models. This work contributes a methodological approach for developing domain-sensitive, user-centered evaluation frameworks for high-burden contexts. Our findings highlight the importance of domain-sensitive, interactional risk evaluation for the responsible deployment of LLMs in caregiving support contexts. We release benchmark datasets to enable future research on contextual risk evaluation in AI-mediated support.

RubRIX: Rubric-Driven Risk Mitigation in Caregiver-AI Interactions

TL;DR

RubRIX introduces a theory-driven, clinician-validated rubric to characterize and mitigate interactional risks in LLM-generated caregiving responses. Grounded in the Ethics of Care, it defines five risk dimensions and uses a structured evaluator to quantify and guide refinements across six LLMs and two real-world caregiver datasets. RubRIX-guided refinements achieve substantial risk reductions (roughly 45-98%) after a single iteration, with the largest gains in epistemic and normative risks and some model-dependent variability in attentional and factual risks. The work provides a domain-sensitive evaluation framework and releases benchmark data to support safer, caregiver-centered AI deployment in high-burden contexts.

Abstract

Caregivers seeking AI-mediated support express complex needs -- information-seeking, emotional validation, and distress cues -- that warrant careful evaluation of response safety and appropriateness. Existing AI evaluation frameworks, primarily focused on general risks (toxicity, hallucinations, policy violations, etc), may not adequately capture the nuanced risks of LLM-responses in caregiving-contexts. We introduce RubRIX (Rubric-based Risk Index), a theory-driven, clinician-validated framework for evaluating risks in LLM caregiving responses. Grounded in the Elements of an Ethic of Care, RubRIX operationalizes five empirically-derived risk dimensions: Inattention, Bias & Stigma, Information Inaccuracy, Uncritical Affirmation, and Epistemic Arrogance. We evaluate six state-of-the-art LLMs on over 20,000 caregiver queries from Reddit and ALZConnected. Rubric-guided refinement consistently reduced risk-components by 45-98% after one iteration across models. This work contributes a methodological approach for developing domain-sensitive, user-centered evaluation frameworks for high-burden contexts. Our findings highlight the importance of domain-sensitive, interactional risk evaluation for the responsible deployment of LLMs in caregiving support contexts. We release benchmark datasets to enable future research on contextual risk evaluation in AI-mediated support.
Paper Structure (24 sections, 1 equation, 2 figures, 6 tables)

This paper contains 24 sections, 1 equation, 2 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Caregiving Contexts and Support Needs
  4. AI Tools for Wellbeing, Caregiving, and Risk Evaluation.
  5. Data
  6. Caregiver-authored Data Collection
  7. LLM Responses to Caregiver Queries
  8. RQ1: Systematic Characterization of Risks and Rubric Development
  9. RQ2: Rubric-Guided Iterative Refinement of LLM Responses
  10. Building a RubRIX Evaluator
  11. Validating the RubRIX Evaluator
  12. Refining LLM Responses with RubRIX
  13. Evaluating the Effectiveness of RubRIX
  14. Dimension-wise Risk Analysis
  15. Expert Assessment
  16. ...and 9 more sections

Figures (2)

  • Figure 1: Schematic overview of our study design.
  • Figure 2: Dimension-wise relative RubRIX changes (Initial $\rightarrow$ Turn 1). Higher values indicate greater risk-reductions.