OmniGuard: Unified Omni-Modal Guardrails with Deliberate Reasoning

TL;DR

OmniGuard introduces a unified omni-modal safety framework with deliberate reasoning, trained on a large diverse dataset, to improve robustness and effectiveness in safeguarding across all modalities in human-AI interactions.

Contribution

The paper presents OmniGuard, the first omni-modal guardrail system capable of safeguarding across all modalities with deliberate reasoning, supported by a large, annotated omni-modal safety dataset.

Findings

Achieves strong safety performance across 15 benchmarks.

Demonstrates robustness in diverse multimodal safety scenarios.

Provides a unified framework for omni-modal risk mitigation.

Abstract

Omni-modal Large Language Models (OLLMs) that process text, images, videos, and audio introduce new challenges for safety and value guardrails in human-AI interaction. Prior guardrail research largely targets unimodal settings and typically frames safeguarding as binary classification, which limits robustness across diverse modalities and tasks. To address this gap, we propose OmniGuard, the first family of omni-modal guardrails that performs safeguarding across all modalities with deliberate reasoning ability. To support the training of OMNIGUARD, we curate a large, comprehensive omni-modal safety dataset comprising over 210K diverse samples, with inputs that cover all modalities through both unimodal and cross-modal samples. Each sample is annotated with structured safety labels and carefully curated safety critiques from expert models through targeted distillation. Extensive experiments on 15 benchmarks show that OmniGuard achieves strong effectiveness and generalization across a wide range of multimodal safety scenarios. Importantly, OmniGuard provides a unified framework that enforces policies and mitigates risks in omni-modalities, paving the way toward building more robust and capable omnimodal safeguarding systems.