Many Ways to Be Fake: Benchmarking Fake News Detection Under Strategy-Driven AI Generation

arXiv cs.CL / 4/13/2026

📰 NewsSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper argues that fake news is increasingly produced via human–AI collaboration, embedding subtle inaccuracies within otherwise credible narratives rather than relying on fully fabricated content.
It introduces MANYFAKE, a synthetic benchmark of 6,798 fake news articles generated with multiple strategy-driven prompting pipelines to reflect diverse real-world deception patterns.
Experiments evaluating state-of-the-art fake news detectors find that advanced reasoning-enabled models can near-saturate on fully fabricated stories.
However, detectors show brittleness against subtler, optimized falsehoods that are interwoven with accurate information, indicating persistent weaknesses in current approaches.
The work highlights a benchmarking gap: mixed-truth cases remain underrepresented, and improved benchmarks are needed to measure robustness against strategy-driven AI generation.

Abstract

Recent advances in large language models (LLMs) have enabled the large-scale generation of highly fluent and deceptive news-like content. While prior work has often treated fake news detection as a binary classification problem, modern fake news increasingly arises through human-AI collaboration, where strategic inaccuracies are embedded within otherwise accurate and credible narratives. These mixed-truth cases represent a realistic and consequential threat, yet they remain underrepresented in existing benchmarks. To address this gap, we introduce MANYFAKE, a synthetic benchmark containing 6,798 fake news articles generated through multiple strategy-driven prompting pipelines that capture many ways fake news can be constructed and refined. Using this benchmark, we evaluate a range of state-of-the-art fake news detectors. Our results show that even advanced reasoning-enabled models approach saturation on fully fabricated stories, but remain brittle when falsehoods are subtle, optimized, and interwoven with accurate information.