AprielGuard

TL;DR

AprielGuard is an 8B parameter unified safeguard model for large language models that detects unsafe and adversarial behaviors across various scenarios, improving robustness and interpretability.

Contribution

It introduces a unified taxonomy and learning framework for safety and adversarial threats, trained on diverse data, with enhanced interpretability and superior performance.

Findings

Outperforms existing guardrails like Llama-Guard and Granite Guardian

Effective in multi-step and reasoning-intensive scenarios

Achieves strong detection of harmful content and adversarial manipulations

Abstract

Safeguarding large language models (LLMs) against unsafe or adversarial behavior is critical as they are increasingly deployed in conversational and agentic settings. Existing moderation tools often treat safety risks (e.g. toxicity, bias) and adversarial threats (e.g. prompt injections, jailbreaks) as separate problems, limiting their robustness and generalizability. We introduce AprielGuard, an 8B parameter safeguard model that unify these dimensions within a single taxonomy and learning framework. AprielGuard is trained on a diverse mix of open and synthetic data covering standalone prompts, multi-turn conversations, and agentic workflows, augmented with structured reasoning traces to improve interpretability. Across multiple public and proprietary benchmarks, AprielGuard achieves strong performance in detecting harmful content and adversarial manipulations, outperforming existing opensource guardrails such as Llama-Guard and Granite Guardian, particularly in multi-step and reasoning intensive scenarios. By releasing the model, we aim to advance transparent and reproducible research on reliable safeguards for LLMs.