Why people trust or distrust AI: Role of emotional response

The growing integration of artificial intelligence (AI) into everyday decision-making has intensified the debate over how people evaluate the reliability and risks of algorithmic systems. While much attention has been given to technical accuracy and data quality, researchers increasingly argue that emotional reactions may significantly shape how individuals interpret and trust AI technologies.

This perspective is explored in the study "Do emotions matter in AI? The mediating role of emotional response between perceived risk and trust," published in AI & Society, which examines the psychological mechanisms linking perceived risk, emotional response, and trust in AI systems.

Emotional responses shape trust in AI systems

Emotional responses act as a mediator between perceived risk and trust in AI. When individuals believe that an AI system poses significant risks, they tend to experience negative emotional reactions, which in turn reduce their willingness to trust the system. On the other hand, when emotional reactions are more positive, individuals are more likely to express confidence in AI decision-making.

This relationship highlights the importance of emotional processing in human–AI interaction. Although AI systems are often designed to operate through objective data analysis, human responses to these technologies are deeply influenced by psychological factors.

The researchers found that perceived risk consistently reduces trust in AI systems, but the mechanism through which this occurs often involves emotional reactions rather than purely cognitive reasoning. Negative emotional responses such as discomfort or unease can intensify concerns about algorithmic decisions, leading individuals to become more skeptical about relying on AI outputs.

Positive emotional responses can mitigate the impact of perceived risk. When individuals feel more comfortable with an AI system, they may still recognize potential risks but remain willing to trust its decisions. This finding suggests that emotional experience plays a key role in shaping public acceptance of AI technologies.

The study also reveals that emotional mediation varies depending on the context in which AI systems operate. In scenarios involving lower levels of automation and lower levels of societal impact, emotional responses can fully explain the link between perceived risk and trust. In other words, individuals' emotional reactions largely determine whether they trust the system.

However, as the stakes of AI decision-making increase, emotional responses become only part of the equation. In high-impact scenarios involving greater automation or more critical consequences, emotional reactions still influence trust but are accompanied by additional considerations related to risk evaluation and accountability.

Cultural differences influence human–AI trust dynamics

The study also sheds light on the role of cultural context in shaping emotional responses to artificial intelligence. The research compares participants from the United Kingdom and the Arab Gulf region to examine whether cultural differences influence how individuals interpret risk and trust in AI systems.

The results show that emotional mediation occurs in both cultural groups, but the strength of this relationship varies across contexts. In the United Kingdom sample, emotional responses played an especially strong role in shaping trust in lower-risk AI scenarios. In these cases, emotional reactions fully explained how perceived risk translated into trust levels.

On the other hand, participants from the Arab Gulf region showed partial emotional mediation across all scenarios, indicating that emotional responses consistently influenced trust but did not entirely account for the relationship between risk and trust.

These differences highlight how cultural factors may influence attitudes toward technology. Societies vary in their levels of technological familiarity, institutional trust, and risk tolerance, all of which can shape how individuals interpret AI systems and their potential consequences.

The study suggests that policymakers and developers should consider cultural context when designing AI governance strategies. Trust in AI is not determined solely by technical performance but also by the social environments in which these technologies are deployed.

Understanding cultural variations in emotional responses may therefore be essential for developing AI systems that are accepted across different regions and communities.

Implications for AI governance and responsible design

If emotional responses play a critical role in shaping trust, then building reliable AI systems requires more than improving algorithmic accuracy. Developers and policymakers must also consider how users emotionally experience interactions with intelligent technologies.

One implication is that transparency and explainability may help reduce negative emotional reactions associated with perceived risk. When users understand how AI systems arrive at decisions, they may feel more comfortable trusting algorithmic outputs even in high-stakes situations.

Communication strategies may also influence emotional responses. Clear explanations about the capabilities and limitations of AI systems can help prevent unrealistic expectations or exaggerated fears. By fostering informed understanding, organizations can reduce emotional uncertainty surrounding algorithmic decision-making.

Another important consideration involves trust calibration, the process of aligning human trust with the actual reliability of AI systems. Both excessive trust and excessive skepticism can lead to problems. Over-reliance on AI may cause users to accept flawed recommendations without scrutiny, while under-trust may prevent organizations from benefiting from AI's analytical capabilities.

The study suggests that emotional responses influence whether individuals achieve this balance. Positive emotional experiences may encourage appropriate reliance on AI systems, while negative emotional reactions can lead to premature rejection of algorithmic tools.

These insights call for interdisciplinary approaches to AI development that integrate technical expertise with insights from psychology, sociology, and behavioral science. Designing AI systems that people trust requires understanding how humans interpret technological risks and how emotional experiences shape decision-making.

The research also highlights the importance of responsible AI governance frameworks that account for human factors alongside technical standards. Regulators and organizations may need to consider how AI systems affect users' emotional perceptions, particularly in sectors where decisions carry significant consequences, such as healthcare, finance, and public administration.