Deceptive Patterns
‹ All reading

Investigating the Impact of Dark Patterns on LLM-Based Web Agents

Author
Devin Ersoy, Brandon Lee, Ananth Shreekumar, Arjun Arunasalam, Muhammad Ibrahim, Antonio Bianchi, Z. B. C. P. University, Florida International University, G. I. O. Technology
Date
20 Oct 2025
Publisher
IEEE Symposium on Security and Privacy
Focus
AI & Automation
Category
Academic Scholar

LiteAgent is introduced, a lightweight framework that automatically prompts agents to execute tasks while capturing comprehensive logs and screen-recordings of their interactions and TrickyArena, a controlled environment comprising web applications from domains such as e-commerce, streaming services, and news platforms, each containing diverse and realistic dark patterns that can be selectively enabled or disabled.

As users increasingly turn to large language model (LLM) based web agents to automate online tasks, agents may encounter dark patterns: deceptive user interface designs that manipulate users into making unintended decisions. Although dark patterns primarily target human users, their potentially harmful impacts on LLM-based generalist web agents remain unexplored. In this paper, we present the first study that investigates the impact of dark patterns on the decisionmaking process of LLM-based generalist web agents. To achieve this, we introduce LiteAgent, a lightweight framework that automatically prompts agents to execute tasks while capturing comprehensive logs and screen-recordings of their interactions.