Although artificial intelligence (AI) shows growing promise for mental health care, current approaches to evaluating AI tools in this domain remain fragmented and poorly aligned with clinical practice, social context, and first-hand user experience. This paper argues for a rethinking of responsible evaluation -- what is measured, by whom, and for what purpose -- by introducing an interdisciplinary framework that integrates clinical soundness, social context, and equity, providing a structured basis for evaluation. Through an analysis of 135 recent *CL publications, we identify recurring limitations, including over-reliance on generic metrics that do not capture clinical validity, therapeutic appropriateness, or user experience, limited participation from mental health professionals, and insufficient attention to safety and equity. To address these gaps, we propose a taxonomy of AI mental health support types -- assessment-, intervention-, and information synthesis-oriented -- each with distinct risks and evaluative requirements, and illustrate its use through case studies.
@article{arxiv.2602.00065,
title = {Responsible Evaluation of AI for Mental Health},
author = {Hiba Arnaout and Anmol Goel and H. Andrew Schwartz and Steffen T. Eberhardt and Dana Atzil-Slonim and Gavin Doherty and Brian Schwartz and Wolfgang Lutz and Tim Althoff and Munmun De Choudhury and Hamidreza Jamalabadi and Raj Sanjay Shah and Flor Miriam Plaza-del-Arco and Dirk Hovy and Maria Liakata and Iryna Gurevych},
journal= {arXiv preprint arXiv:2602.00065},
year = {2026}
}