CHASM: Unveiling Covert Advertisements on Chinese Social Media

arXiv cs.LG / 4/23/2026

📰 NewsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper highlights a gap in existing LLM/social-media moderation benchmarks: they often miss covert advertisements that impersonate normal posts to manipulate consumer purchases.
Researchers introduce CHASM, a manually curated, anonymized multimodal dataset (4,992 instances) drawn from real scenarios on China’s Rednote platform, created with strict privacy and quality controls.
Evaluation results show that in both zero-shot and in-context learning modes, today’s multimodal LLMs are not reliably able to detect covert ads.
While fine-tuning open-source MLLMs on CHASM improves performance, the study finds persistent difficulty in spotting subtle cues in comments and in distinguishing nuanced visual vs. textual patterns.
The authors provide detailed error analysis and call for better defenses from the research community and social-media moderators against this emerging threat.

Abstract

Current benchmarks for evaluating large language models (LLMs) in social media moderation completely overlook a serious threat: covert advertisements, which disguise themselves as regular posts to deceive and mislead consumers into making purchases, leading to significant ethical and legal concerns. In this paper, we present the CHASM, a first-of-its-kind dataset designed to evaluate the capability of Multimodal Large Language Models (MLLMs) in detecting covert advertisements on social media. CHASM is a high-quality, anonymized, manually curated dataset consisting of 4,992 instances, based on real-world scenarios from the Chinese social media platform Rednote. The dataset was collected and annotated under strict privacy protection and quality control protocols. It includes many product experience sharing posts that closely resemble covert advertisements, making the dataset particularly challenging.The results show that under both zero-shot and in-context learning settings, none of the current MLLMs are sufficiently reliable for detecting covert advertisements.Our further experiments revealed that fine-tuning open-source MLLMs on our dataset yielded noticeable performance gains. However, significant challenges persist, such as detecting subtle cues in comments and differences in visual and textual structures.We provide in-depth error analysis and outline future research directions. We hope our study can serve as a call for the research community and platform moderators to develop more precise defenses against this emerging threat.