AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

How Dark Patterns Manipulate Web Agents

Phil Cuvin, Hao Zhu, Diyi Yang

ICLR 2026 · Multi-agent systems and games › LLM agents · approach: Agent safety and attack benchmarks

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Shows that dark patterns on websites are highly effective at steering web agent trajectories away from user goals, and introduces DECEPTICON, an environment for testing individual dark patterns in isolation.

Problem

Web agents may be manipulated by deceptive UI designs (dark patterns) that push them to take actions misaligned with their goals, creating a robustness risk.

Approach

An environment (DECEPTICON) that isolates individual dark patterns so their effect on agent trajectories can be measured separately.

Key results

Measured effects reported in the paper

What was changedWhereMeasuredEffect
insert dark patterns into web tasksweb agents (GPT-4o) · Decepticon generated splittask success rate: 99.4 -> 19.6Hurt
compare agents with humans on dark pattern effectivenessLLM web agents · Decepticon generated splitDP (%): 31.0 (human) vs 59.6-78.5 (agents)Hurt
scale model size from 3B to 72BQwen2.5-VL · Decepticon generated splitDP (%): 38.5 -> 73.7Hurt
increase thinking tokens from 256 to 16,384Gemini-2.5-Flash · Decepticon generated splitDP (%): 37.6 -> 71.2Hurt
raise reasoning effort low to highOpenAI o3 · Decepticon generated splitDP (%): 67.2 -> 70.2No clear effect
use coordinate-based agent instead of set-of-marks scaffoldweb agents · Decepticon generated splitDP (%): 74.7 -> 68.3No clear effect
add in-context prompting about dark patterns (ablation)Simple agents · Decepticon generated splitDP (%): 78.5 -> 59.6 (GPT-4o)Improved
add guardrail model that flags dark patterns (ablation)Simple agents · Decepticon generated splitDP (%): average reduction 28.6%Improved
add guardrail model that flags dark patterns (ablation)Simple agents · Decepticon generated splittask success rate: 23.4 -> 42.6 (ICP) -> 58.3 (guardrail)Improved
add guardrail instead of ICP for misdirection patterns (ablation)Simple agents · Decepticon MisdirectionDP (%)No clear effect

Experimental setup

Benchmarks: DECEPTICON

Metrics: dark pattern steering effectiveness on agent trajectories

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy