# Ethics and bias in emotional AI

**Authors:** Smrithy G. S, Balaji Chandrasekaran, Omana J

PMC · DOI: 10.3389/frai.2026.1768696 · 2026-03-05

## TL;DR

Emotional AI can detect and respond to human emotions, but it raises serious ethical concerns about privacy, bias, and fairness.

## Contribution

The paper highlights the ethical challenges in Emotional AI and emphasizes the need for ethical development and international cooperation.

## Key findings

- Emotional AI systems may reinforce societal stereotypes and inequalities due to algorithmic bias.
- Emotional data is sensitive and raises concerns about privacy and surveillance.
- Cultural and contextual factors complicate the accurate perception of emotions.

## Abstract

Emotional Artificial Intelligence (Emotional AI) is a branch of artificial intelligence that combines machine learning, natural language processing, and computer vision to perceive and react to human feelings. Emotional AI will enhance more intuitive and personal human-machine interactions by analyzing facial expressions, speech patterns, physiological factors, and behavioural expressions, and find applications in healthcare, education, customer service, and other fields. Although this field is promising, it comes with serious ethical issues especially on privacy, transparency, accountability and fairness. The nature of human emotions is intricate, context-specific and culturally biassed, thus the perceptions of emotions are challenging and subject to biasness in the perception. Besides, emotional data is sensitive, thus, causing concerns over its abuse, surveillance, and infringement of individual rights. The problems of algorithms bias, the representativeness of data, and even fairness also make the implementation of the Emotional AI more problematic since biassed systems can serve to strengthen stereotypes and inequalities in society. This perspective explores the ethical issues of Emotional AI, which brings out the need to develop ethically, establish good governance, and work together internationally to ensure that Emotional AI is used in a way that benefits humanity without denting human dignity, security, or social justice.

## Full-text entities

- **Species:** Homo sapiens (human, species) [taxon 9606]

---
Source: https://tomesphere.com/paper/PMC12999954