AI Models Surpass Humans in Emotional Intelligence Tests

This updated guide explores AI models surpass humans in emotional intelligence tests with practical details, clear explanations, and useful takeaways. Modern artificial intelligence (AI) continues to amaze us with its capabilities beyond expectations, especially large language models (LLMs). Tech giants like OpenAI, DeepSeek, Google, and Meta are working on a series of more efficient AI models, expanding their potential.

AI Understands Emotions Better Than Humans

A team of researchers from the University of Geneva and the University of Bern (Switzerland) published their groundbreaking findings in the journal Communications Psychology . The study compared six leading LLMs — including ChatGPT-4, Gemini 1.5 Flash, Claude 3.5 Haiku, Copilot 365, and DeepSeek V3 — to humans on five standardized EI tests (a series of multiple-choice questions designed to assess your emotional intelligence), assessing these aspects:

1. Ability to understand emotions
2. Emotional regulation
3. Emotional Management

The test scenarios were designed to be realistic and contain complex emotional elements. The results showed that all AI models outperformed human participants in analyzing and responding based on emotional context.

To test its reliability, the team asked ChatGPT-4 to generate new EI scenarios on its own. These scenarios were then rated by human participants — and the results remained the same: the AI maintained a high level of emotional understanding and reasoning.

Professor Marcello Mortillaro (research team member) commented:

LLMs not only select the optimal answer from the available choices, but also generate new scenarios that fit the context. This strengthens the hypothesis that LLMs like ChatGPT possess emotional knowledge and are capable of emotional reasoning .

This discovery opens the door to applying AI in areas that require emotional sophistication, previously thought to be the 'exclusive' domain of humans:

• Coaching: AI can provide constructive feedback, helping improve communication skills.
• Education: Support to adjust teaching methods based on students' emotional state.
• Conflict Management: Analyze conflicts and propose amicable solutions.

Despite the impressive results, experts caution that AI's ability to truly sense emotions remains an open question. Passing the test does not necessarily mean that AI can 'feel' emotions like humans. But the study does confirm that AI is getting closer to emulating emotional understanding — a key element in human-machine interactions.