AprielGuard

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.

Key Contributions

• Unified taxonomy covering both safety risks (toxicity, bias) and adversarial threats (prompt injection, jailbreaks) across standalone, conversational, and agentic interaction modes
• AprielGuard: an 8B safeguard model trained on diverse open and synthetic data with structured reasoning traces for interpretability
• First open-source guardrail with explicit agentic jailbreak detection, outperforming Llama-Guard and Granite Guardian on multi-step and reasoning-intensive benchmarks

🛡️ Threat Analysis

Details

Similar Papers