LLM App Squatting and Cloning

TL;DR

This study presents the first large-scale analysis of LLM app squatting and cloning using its custom-built tool, LLMappCrazy, which integrates Levenshtein distance and BERT-based semantic analysis to detect cloning by analyzing app functional similarities.

Abstract

Impersonation tactics, such as app squatting and app cloning, have posed longstanding challenges in mobile app stores, where malicious actors exploit the names and reputations of popular apps to deceive users. With the rapid growth of Large Language Model (LLM) stores like GPT Store and FlowGPT, these issues have similarly surfaced, threatening the integrity of the LLM app ecosystem. In this study, we present the first large-scale analysis of LLM app squatting and cloning using our custom-built tool, LLMappCrazy. LLMappCrazy covers 14 squatting generation techniques and integrates Levenshtein distance and BERT-based semantic analysis to detect cloning by analyzing app functional similarities. Using this tool, we generated variations of the top 1000 app names and found over 5,000 squatting apps in the dataset. Additionally, we observed 3,509 squatting apps and 9,575 cloning cases across six major platforms. After sampling, we find that 18.7% of the squatting apps and 4.9% of the cloning apps exhibited malicious behavior, including phishing, malware distribution, fake content dissemination, and aggressive ad injection.

Figures (11)

• Figure 1: An example of LLM app squatting.
• Figure 2: Our approach to identifying squatting and cloning LLM apps.
• Figure 3: The 14 kinds of LLM app squatting-generation models used in this work. The 6 models in black are inherited from AppCrazy hu2020mobile, while the 8 models in red that are either newly introduced or modified in LLMappCrazy to target LLM apps.
• Figure 4: The distribution of squatting apps across models.
• Figure 5: A real-world example highlights the differences between Levenshtein and BERT-based semantic similarity methods. Although all three apps convey the same core meaning, a typographical error with the term "fasten" in App2 and App3 causes the Levenshtein method to detect similarity only between these two, missing the similarity between App1 and App2.