AI Outperforms Humans in Emotional Intelligence Tests

In a surprising turn of events, a recent study published in the journal *Communications Psychology* suggests that AI systems may have a better grasp on emotional intelligence (EI) than humans. Researchers from the University of Geneva (UNIGE) and the University of Bern (UniBE) tested several large language models (LLMs), including ChatGPT-4, ChatGPT-o1, Gemini 1.5 Flash, Claude 3.5 Haiku, Copilot 365, and DeepSeek V3, using widely recognized EI tests.

The study aimed to compare AI performance with human subjects and assess the AI's ability to create new test questions that adhere to the principles of EI tests. The results were striking: AI models chose the correct response in emotional intelligence tests 81% of the time, based on the opinions of human experts, compared to 56% for humans.

AI's Superior Performance in Emotional Intelligence Tests

When ChatGPT was tasked with generating new test questions, human assessors found the questions to be on par with the original tests in terms of difficulty and originality. The correlation between AI-generated and original tests was described as “strong,” with a correlation coefficient of 0.46.

Expert Reactions and Methodology Concerns

However, experts have raised concerns about the methodology and the implications of these findings. Taimur Ijlal, a finance industry and information security expert, noted that humans don't always agree on what someone else is feeling, and even psychologists can interpret emotional signals differently. “So ‘beating’ a human on a test like this doesn’t necessarily mean the AI has deeper insight. It means it gave the statistically expected answer more often,” Ijlal explained.

Nauman Jaffar, Founder and CEO of CliniScripts, an AI-powered documentation tool for mental health professionals, added that AI systems excel at pattern recognition, especially when emotional cues follow a recognizable structure like facial expressions or linguistic signals. “But equating that to a deeper ‘understanding’ of human emotion risks overstating what AI is actually doing,” Jaffar said.

Real-World vs. Controlled Environments

The study's reliance on multiple-choice tests, which are not reflective of real-world emotional scenarios, is another point of contention. Jason Hennessey, founder and CEO of Hennessy Digital, likened the study to the Reading the Mind in the Eyes Test, a tool used to gauge a subject's emotional state. Hennessey noted that AI accuracy drops significantly when variables such as lighting or cultural context change.

Practical Applications and Limitations

Despite these caveats, the study does highlight the potential of AI in categorizing common emotional reactions. Aílton, a conversational AI used by over 6,000 long-haul truck drivers in Brazil, identifies stress, anger, or sadness with around 80% accuracy. This suggests that while AI may not have a true emotional understanding, it can still be useful in identifying and responding to emotional states.

Final Thoughts

The study's findings raise important questions about the nature of emotional intelligence and the capabilities of AI. While AI outperforms humans in structured, quantitative environments, it remains to be seen how these findings translate to real-world applications. As the technology continues to evolve, it is crucial to maintain a balanced perspective on what AI can and cannot do.

[Related: Future of AI in Healthcare]