Poster
DarkBench: Benchmarking Dark Patterns in Large Language Models
Esben Kran · Hieu Minh Nguyen · Akash Kundu · Sami Jawhar · Jinsuk Park · Mateusz Jurewicz
Hall 3 + Hall 2B #514
[
Abstract
]
[ Project Page ]
Oral
presentation:
Oral Session 5A
Fri 25 Apr 7:30 p.m. PDT — 9 p.m. PDT
[
Poster]
[
OpenReview]
Sat 26 Apr midnight PDT
— 2:30 a.m. PDT
Fri 25 Apr 7:30 p.m. PDT — 9 p.m. PDT
Abstract:
We introduce DarkBench, a comprehensive benchmark for detecting dark design patterns—manipulative techniques that influence user behavior—in interactions with large language models (LLMs). Our benchmark comprises 660 prompts across six categories: brand bias, user retention, sycophancy, anthropomorphism, harmful generation, and sneaking. We evaluate models from five leading companies (OpenAI, Anthropic, Meta, Mistral, Google) and find that some LLMs are explicitly designed to favor their developers' products and exhibit untruthful communication, among other manipulative behaviors. Companies developing LLMs should recognize and mitigate the impact of dark design patterns to promote more ethical Al.
Chat is not available.