The Body-Algorithm-Emotion Loop: Making Emotion AI Misrecognition Visible through Artistic Practice

Emotion AI systems frequently misclassify facial expressions, and commercial systems translate facial movements into emotional categories that may not accurately reflect what people actually feel—a problem the artwork makes visible by letting viewers experience real-time misrecognition of their own faces.

Algorithmic classification can enforce compliance, as Emotion AI doesn't just describe emotions but judges whether people meet standards of "compliance or non-compliance," raising concerns about how these systems might be used to monitor and control behavior in workplaces, schools, or other institutions.

Explainability requires more than technical transparency—understanding how AI systems work isn't enough; people need embodied, experiential encounters with how algorithms measure and classify their bodies to truly grasp the stakes of these technologies.

Artistic practice offers a valuable method for AI accountability by making algorithmic processes visible and contestable in ways that technical documentation alone cannot, allowing people to directly experience and question how their bodies are being interpreted by machines.

Emotion AI reflects and reinforces power imbalances, particularly through what the work calls "algorithmic colonialism"—the imposition of standardized emotional categories that may not reflect diverse cultural expressions, body types, or ways of being.