RADAR: Retrieval-Augmented Detector with Adversarial Refinement for Robust Fake News Detection

TL;DR

RADAR is a novel fake news detection system that combines retrieval-augmented verification with adversarial refinement, using structured natural language critiques to enhance robustness against sophisticated misinformation tactics.

Contribution

The paper introduces VAF, a structured adversarial feedback mechanism, and demonstrates how retrieval-augmented detection with adversarial refinement improves fake news detection robustness.

Findings

RADAR outperforms existing baselines on benchmark datasets.

Detector-side retrieval significantly boosts detection accuracy.

VAF and few-shot demonstrations provide complementary robustness benefits.

Abstract

To efficiently combat the spread of LLM-generated misinformation, we present RADAR, a Retrieval-Augmented Detector with Adversarial Refinement for robust fake news detection. Our approach employs a generator that rewrites real articles with factual perturbations, paired with a lightweight detector that verifies claims using dense passage retrieval. To enable effective co-evolution, we introduce verbal adversarial feedback (VAF). Rather than relying on scalar rewards, VAF issues structured natural-language critiques; these guide the generator toward more sophisticated evasion attempts, compelling the detector to adapt and improve. On a fake news detection benchmark, RADAR consistently outperforms strong retrieval-augmented trainable baselines, as well as general-purpose LLMs with retrieval. Further analysis shows that detector-side retrieval yields the largest gains, while VAF and few-shot demonstrations provide complementary benefits. RADAR also transfers better to fake news generated by an unseen external attacker, indicating improved robustness beyond the co-evolved training setting.