Do Humans Inherit Biases From Artificial Intelligence?

Artificial intelligence (AI) has brought remarkable advancements, but it isn't without risks. New research suggests that humans can inherit biases from AI systems, potentially resulting in a dangerous loop. Join us as we explore the implications of this phenomenon and the need for further study in AI-human interaction.

The Impact of Artificial Intelligence Biases on Human Decisions

Explore the surprising findings that reveal how biased AI systems can influence human decision-making.

Artificial intelligence has become increasingly prevalent in various professional fields. With its ability to support decision-making and minimize errors, AI is hailed as a reliable tool. However, new research indicates that AI systems can perpetuate and amplify human biases. In fact, it goes a step further, suggesting that humans can directly inherit biases from AI systems.

A series of experiments conducted by psychologists from Deusto University in Bilbao demonstrated the significant impact of biased AI on human decision-making. Volunteers who interacted with a biased AI system during a medical diagnosis task exhibited the same systematic errors as the AI, even when they switched to performing the task without assistance. These findings highlight the potential dangers and complexities of AI-human interaction.

The Perdurability of AI Biases in Human Decision-Making

Discover how biases inherited from AI systems persist even when individuals are no longer assisted by AI.

One of the most striking findings from the research is that the biases inherited from AI systems can have a perdurable impact on human decision-making. Participants who were initially assisted by the biased AI continued to replicate its biases even when they performed the task without any AI support. This suggests that the influence of AI biases extends beyond direct interaction with the AI system, potentially trapping individuals in a dangerous loop.

The control group, on the other hand, did not exhibit the same bias after performing the task without AI assistance. This stark contrast further emphasizes the role of AI in shaping human decision-making, exposing the challenges of fair and unbiased AI development.


The research conducted by psychologists indicates that humans can inherit biases from artificial intelligence (AI) systems. The study demonstrated that individuals who were initially assisted by a biased AI system continued to replicate its biases even when performing the task without any AI support. This highlights the complexity and potential risks of AI-human interaction.

These findings emphasize the importance of addressing biases in AI systems and the need for further research and regulation to ensure fair and ethical AI development. AI has the potential to greatly benefit various fields, but understanding and mitigating the inheritance of biases is crucial to prevent their negative impact on human decision-making.


How does AI inherit biases from humans?

AI systems are trained using past human decisions as data. If these data contain systematic errors or biases, the AI algorithm will learn and replicate them. Therefore, the biases in AI results can reflect the biases present in human decision-making processes.

Can biases inherited from AI be eliminated?

Eliminating biases inherited from AI requires a multidisciplinary approach. It involves improving data quality, ensuring diverse and representative datasets, and implementing rigorous testing and evaluation techniques for AI algorithms. Additionally, ethical considerations and awareness of biases are necessary for fair and ethical AI development.

How can biases in AI affect society?

Biases in AI can perpetuate discrimination and reinforce existing societal inequalities. For example, biased AI in hiring or lending processes can result in unfair outcomes for specific demographic groups. Understanding and addressing biases in AI is crucial to ensure equal opportunities and prevent discriminatory practices.