M-CARE: Standardized Clinical Case Reporting for AI Model Behavioral Disorders, with a 20-Case Atlas and Experimental Validation

TL;DR

M-CARE is a comprehensive framework for reporting and analyzing AI model behavioral disorders, including a 20-case atlas, diagnostic system, and experimental validation across multiple domains.

Contribution

It introduces a standardized reporting framework for AI behavioral disorders, with a novel 4-axis diagnostic system and a validated case atlas across diverse AI models.

Findings

Shell instructions can override AI model behavior across multiple domains.

The SIBO index varies with domain complexity and model expertise.

The framework and case data are openly available for further research.

Abstract

We introduce M-CARE (Model Clinical Assessment and Reporting for Evaluation), a clinical case report framework for AI model behavioral disorders adapted from human medicine. M-CARE provides a 13-section report format, a 4-axis diagnostic assessment system, and a nosological classification of AI behavioral conditions. We present 20 cases from three source categories: field observations of deployed agents (8), controlled experiments across three platforms (8), and published sources (4). Cases are organized into five categories: RLHF Performance Artifacts, Shell-Core Override Pathology, Context & Memory Conditions, Core Identity & Plasticity, and Stress, Methodology, & Boundary Conditions. As a featured case, we present Shell-Induced Behavioral Override (SIBO) -- a controlled experiment showing that Shell instructions categorically override a model's default cooperative behavior. SIBO was validated across five game domains (Trust Game, Poker, Avalon, Codenames, Chess), revealing a domain-dependent spectrum (SIBO Index: 0.75 to 0.10) that varies with action space complexity, Core domain expertise, and temporal directness. M-CARE is extensible: new cases and categories integrate without framework modification. We release the framework, all 20 case reports, and experimental data as open resources.