The AI Accomplice

DEADLINE: February 08, 2026 AoE

AI can empower creativity and amplify manipulation. This workshop brings researchers together to build safer, more transparent AI-generated experiences.

Generative AI is becoming a powerful assistant in design work. Tools like LLMs and image generators can create interfaces, content, and entire digital experiences in seconds. But these models also learn from the web, including harmful design patterns. As a result, AI can introduce or even amplify deceptive elements that manipulate users without their awareness.

This workshop brings together the HCI community to better understand these risks and to design safeguards that keep user autonomy and transparency at the center of AI-generated experiences.

Program

Monday, April 13, 2026 // 14:15 – 15:45 CEST and 16:30 – 18:00 CEST

For our workshop, we follow a two-parted schedule. In the first part we focus on sharing the ideas and concepts, while the second part introduces hands-on approaches to showcase deceptive design and how to counteract them.

14:15 - 14:30 CEST

Opening

14:30 - 15:30 CEST

Presentations & Inspirations for Bad User Interfaces

Generative AI and the Emergence of Deceptive Interaction in Student-Made Net-Art - Afra Pascual, Sergio Sayago, Marc Gonzalez and Alberto Pedrero
AI Iterated Deceptive Patterns - Eszter Vigh
Why Conversational AI is Deceptive Per Se: Critical Reflections On The Appropriate Mental Model For Interacting With LLMs. - Sarah Diefenbach, Daniel Ullrich and Marie Veihelmann
Towards Detecting and Mitigating Deceptive Patterns with GenAI - Niklas Pfützenreuter
Do AI-Enabled Website Builders Dream of Dark-Patterned Sheep? Understanding AI-Enabled Design Tools - Johanna Gunawan, Shashank Chakravarthy and Hannes Westermann
Exploring the “Banality” of Deception in Generative AI - Ishitaa Narwane, Johanna Gunawan and Konrad Kollnig

15:30 - 15:45 CEST

Group Work Preparation

15:45 - 16:30 CEST

Break

16:30 - 16:45 CEST

Group Work: Build a Deceptive Design with AI

16:45 - 17:15 CEST

Group Work: Countering Deceptive Design

17:15 - 17:45 CEST

Group Work: Presentations

17:45 - 18:00 CEST

Closing

Organizers

Thomas Kosch

is an assistant professor at Humboldt University of Berlin, Germany, where he studies how AI can facilitate an understanding between users and interfaces using computational interaction design. Furthermore, his work advances methodological rigor in human–AI interaction research by developing frameworks that improve reproducibility and transparency

Veronika Krauß

is a professor at the University of Applied Sciences Ansbach, Germany. Her work investigates social challenges originating from UI design for and with emerging technologies. Her main focus lies on adapting deceptive designs to XR and the role of design practices and tools.

Christopher Katins

is a PhD student and HCI researcher at the Humboldt University of Berlin. His research focuses on the challenges and dangers that Mixed Reality might pose in the future.

Dominik Schön

is a PhD student at the Technical University of Darmstadt, Germany. In his research, he investigates Ergonomic-aware Extended Reality Environments. In this regard, he examines how AI and LLMs can facilitate the streamlined implementation of such functional design.

Mark McGill

is a senior lecturer (associate professor) in the School of Computing Science, University of Glasgow. His research explores the future of spatial computing and a world where perception is mediated by AI through everyday augmented reality.

Jan Gugenheimer

is a Professor at the Technical University of Darmstadt. His research focuses on the emerging social challenges for mixed reality technology and how to integrate HMDs into the fabric of our daily lives.

© Dominik Schön, Thomas Kosch.
Design: HTML5 UP and a bit of boomer-coding (not vibe-coding)