mlsanityData to 7 Oct 2026

Visual Distraction Undermines Moral Reasoning in Vision-Language Models

arXiv 2603.16445 ↗First posted 17 Mar 2026cs.AIcited by 0

On the map: LLM reasoning → Reasoning benchmarks and evaluation

also in 1 more
Abstract (arXiv)

Moral reasoning is fundamental to safe Artificial Intelligence (AI), yet ensuring its consistency across modalities becomes critical as AI systems evolve from text-based assistants to embodied agents. Current safety techniques demonstrate success in textual contexts, but concerns remain about generalization to visual inputs. Existing moral evaluation benchmarks rely on textonly formats and lack systematic control over variables that influence moral decision-making. Here we show that visual inputs fundamentally alter moral decision-making in state-of-the-art (SOTA) Vision-Language Models (VLMs), bypassing text-based safety mechanisms. We introduce Moral Dilemma Simulation (MDS), a multimodal benchmark grounded in Moral Foundation Theory (MFT) that enables mechanistic analysis through orthogonal manipulation of visual and contextual variables. The evaluation reveals that the vision modality activates intuition-like pathways that override the more deliberate and safer reasoning patterns observed in text-only contexts. These findings expose critical fragilities where language-tuned safety filters fail to constrain visual processing, demonstrating the urgent need for multimodal safety alignment.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 13 of its 63 references that are in the map. First posted on arXiv 17 Mar 2026.

Builds on

13 of 63 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025The Staircase of Ethics: Probing LLM Value Priorities through Multi-Step Induction to Complex Moral Dilemmas23 May

Other references · 12

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2024GPT-4o System Cardcited by 4,467
  4. 2025Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilitiescited by 3,655
  5. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 2,950
  6. 2025Qwen3-VL Technical Reportcited by 2,534
  7. 2025SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Featurescited by 1,127
  8. 2020Aligning AI With Shared Human Valuescited by 706
  9. 2024Yi: Open Foundation Models by 01.AIcited by 706
  10. 2025Generative Value Conflicts Reveal LLM Prioritiescited by 13
  11. 2024MM-MoralBench: A MultiModal Moral Evaluation Benchmark for Large Vision-Language Modelscited by 9
  12. 2026A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5cited by 6

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Do VLMs Have a Moral Backbone? A Study on the Fragile Morality of Vision-Language Models
  2. 2025MORALISE: A Structured Benchmark for Moral Alignment in Visual Language Models
  3. 2026MM-SCALE: Grounded Multimodal Moral Reasoning via Scalar Judgment and Listwise Alignment
  4. 2026Moral Sycophancy in Vision Language Models
  5. 2026SAVeS: Steering Safety Judgments in Vision-Language Models via Semantic Cues
  6. 2025MoralCLIP: Contrastive Alignment of Vision-and-Language Representations with Moral Foundations Theory
  7. 2022When to Make Exceptions: Exploring Language Models as Accounts of Human Moral Judgment
  8. 2026Frame-Conditioned Moral Computation in LLaMA 3.1-8B-Instruct: A Mechanistic Interpretability Audit of Ethical Reasoning