AI Outsmarts Humans in Emotional Intelligence Tests, Say Swiss Researchers

Can AI Outperform Humans in Emotional Intelligence?

Recent research from the University of Geneva (UNIGE) and the University of Bern (UniBE) reveals that generative artificial intelligence (AI) models can outperform humans in emotional intelligence tests. Large language models (LLMs) like ChatGPT scored 82% on these tests, compared to 56% for human participants. This finding challenges assumptions about AI’s capabilities in emotionally charged situations.

Testing Emotional Intelligence with AI

The study involved six generative AI systems, including ChatGPT-4, Gemini 1.5, and Claude 3.5 Haiku. Researchers submitted these models to five established emotional intelligence tests. These tests presented scenarios designed to evaluate skills such as emotional understanding, regulation, and management.

For example, one scenario posed the question: If a colleague steals your idea and receives praise, what’s the most effective response? Options ranged from confronting the colleague to silently resenting them. The AI models consistently chose the most emotionally intelligent option, such as discussing the issue with a supervisor, demonstrating their ability to recommend appropriate responses.

AI’s Edge in Emotional Reasoning

With a significant margin over humans, the tested LLMs demonstrated a deep grasp of social and emotional cues. This suggests that these models have internalized extensive knowledge about human emotions and effective behavior in complex interpersonal situations.

Creating New Emotional Intelligence Tests

Beyond answering existing tests, the researchers tasked ChatGPT-4 with generating new emotional intelligence scenarios. These AI-created tests were then administered to over 400 human participants and showed reliability and clarity comparable to years of expert-developed assessments.

This capability to both evaluate and create emotionally relevant content highlights AI’s potential as a tool in education, coaching, and conflict management — fields where emotional insight is critical.

Implications for Management and Research

• Education and Training: AI can support the development of emotional intelligence curricula and personalized coaching strategies.
• Conflict Resolution: Generative AI could assist in suggesting constructive responses during workplace disputes.
• Psychological Assessment: Automating test generation can speed up research and provide scalable tools for emotional skills evaluation.

These applications require expert supervision to ensure ethical and context-appropriate use. But the results clearly indicate that AI can handle tasks once thought exclusive to human emotional judgment.

For professionals interested in AI’s evolving role in emotional intelligence and management, exploring specialized AI training can provide valuable skills. Check out resources such as latest AI courses to deepen your understanding of practical AI applications.