Deceptive Patterns
‹ All reading

Identifying Dark Patterns in Social Robot Behavior

Author
Elizabeth Dula, Andres Rosero, Elizabeth Phillips
Date
27 Apr 2023
Focus
AI & Automation
Category
Academic Scholar

Social robots are increasingly used in intimate settings like elder care and child education. This raises concerns about deceptive robot design and the potential for dark patterns to exploit vulnerable users by leveraging emotional bonds, as these robots can identify and respond with emotions. This research suggests ways dark patterns could manifest in human-robot relationships and offers ethical design recommendations.

Social robots have become increasingly utilized in intimate environments where their roles can include caretakers for the elderly, general physical or emotional support, entertainment, and educators for children. To accommodate for these increasingly intimate relationships, robotics companies have begun employing robotics with the ability to identity emotions and respond with emotionality in return. This faux emotional relationship opens the door for potential user manipulation and exploitation through deceptive robot design. Dark patterns are deceptive design patterns used by websites or apps to manipulate users into actions the user did not intend. We argue that dark patterns can be programmed into social robotics to leverage these unidirectional human - robot emotional bonds to manipulate users, which could result in the exploitation of vulnerable populations like children and the elderly. Drawing from the dark pattern and social robotics literature, we suggest ways that dark patterns can manifest themselves in these relationships. We also provide recommendations for ethical practices when designing emotional social robots.