Paper page - SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation

Source: https://huggingface.co/papers/2606.03348

Abstract

AI-generated images with realistic text and layouts pose a significant misinformation threat requiring new detection benchmarks and methods beyond surface-level credibility assessment.

Recentgenerative modelscan now produce visual artifacts with realistic embedded text and layouts, creating a newmisinformationthreat:synthetic credibility. We introduceSYNCRED-Bench, a benchmark of 600 AI-generatedmisinformationimages balanced across six credible-form categories and seven fine-grained circulation styles, together withFP450, a real-image negative set for measuring false positives. Extensive evaluation shows that existing systems remain unreliable: under a 5%false-positive-rateconstraint, 15MLLMsachieve only 10.5%true positive rate(TPR), open-sourceAIGC detectorsachieve less than 5%, andcommercial APIsreach 57.6%. Human annotators also struggled to identifysynthetic credibility, reaching only 63% TPR. These findings establishsynthetic credibilityas a severe and underexplored visualmisinformationchallenge, and provide a benchmark for developing detectors that reason beyond superficial credibility cues.

View arXiv pageView PDFAdd to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.03348 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.03348 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.03348 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.