Artificial Intelligence and Cognitive Bias

Author: Evelyn Fistler

Editor: Ellis Mckeown

Introduction

Cognitive biases are systematic and unconscious  errors that impact thinking, occurring when influences such as the environment impact judgement and decision making, often distorting an individual's perception of reality, resulting in misinformation or poor interpretations of information (Da Silva, et al, 2023). Different cognitive biases can impact decision-making and reasoning due to individual beliefs, which can be flawed through influence and reliance on a particular belief, leaving the individual susceptible to flawed reasoning. With the new use of artificial intelligence and chatbots in media platforms and online, people have become more reliant on discussing questions or seeking advice through AI.This article will provide insight into how the modern use of artificial intelligence impacts cognition, potentially affecting critical thinking and reasoning, leading to diminished brain capability.

Types of Cognitive Biases

Familiarity, availability, and confirmation biases are some of the most common cognitive biases that many individuals are vulnerable to. Familiarity bias, a bias where individuals prefer to follow what is already known, suggests that people can be susceptible to false information because of previous set beliefs. Information can spread quickly, whether true or not, and contributes to familiarity bias because of how often the information is read and heard. Believing in the topics and information most easiest to remember is known as availability bias, and because AI provides individuals with quick information that sounds believable because of its delivery, this leaves many relying on false information because of how believable it was initially obtained. 

Utilizing AI, which caters toward “for you” pages and recommendations through shopping and social media apps, can contribute to familiarity and availability bias. Availability bias is a key factor that consumers experience, specifically regarding impulse buying. Frequently encountered products are stored in long term memory, leaving individuals vulnerable to familiarity and therefore a potential purchase (Li, et al, 2025). Because AI has the capability to bring forth information through different platforms while also handling quick assistance, providing consumers with products recommended based on searches and advice or solutions, availability bias can affect individuals based on the easy influx of information.

Through confirmation bias, many seek information that may support personal beliefs and expectations, which reinforces the bias and assumptions. For example, those seeking a health diagnostic based on symptoms may lean toward a self-diagnosis that best describes their personal needs and/or anxieties, leaving them susceptible to confirmation bias. Because AI models have human-like dialogue capabilities and are utilized to cater toward individuals seeking answers, there is a chance for confirmation bias because some individuals may seek health advice. This leaves the individual more vulnerable to misinformation and misinterpretation (Lopez‐Lopez et al., 2025). 

Understanding that AI can easily be used to support personal belief through frequency usage depending on how prompts are worded and understood by the user. Familiarity with AI is suggested to potentially empower scientists to utilize AI visualizations, however, audiences could become further susceptible to misinterpretation and distortion. With this, Eom (et al, 2025), suggests that becoming more familiar with AI and the content generated using different softwares is essential for individuals to become less skeptical. AI disclaimers and credibility along with transparency and understanding for generative AI is extremely important to implement through education and communication to the public.

What Can Influence Cognitive Biases? 

Emotion, personal belief, influenced thoughts and the capability for an individual to process information are all common factors that can influence cognitive bias. Reliance on flawed information, and the perspective of how information is framed, are other considerations for cognitive bias formation, especially with the use of AI (Paulus, et al, 2022). 

The use of artificial intelligence is becoming a part of daily life, whether it is to use AI for ideas, to communicate, or to solve problems with, the human capability to make decisions from AI seems to be causing disruption and affecting cognitive abilities like critical thinking, reasoning, and decision making because humans reproduce the same biases displayed by AI (Vicente & Matute, 2023). Decisions made through biased AI recommendations show how impactful AI is on cognition, with AI responses inheriting the  structure of reasoning and evidence formation that change human behavior. Because AI is man-made, there is room for error especially in generative AI models, where displays of human-like cognitive biases are prevalent (Vicente & Matute, 2023). 

How the Use of Artificial Intelligence Impacts Cognition

AI has the ability to respond quickly because of pre-trained models and pattern recognition through datasets, enabling rapid data processing to summarize data at high speeds.  Promoting cognitive offloading (the act of reducing mental processing requirements), AI responses and continuous usage appears to distort critical thinking, further impacting learning and reasoning, and increases exposure to biased information because Large Language Models (which AI are trained on) can provide narrow ideas that result in superficial analyses (Duke Learning Innovation & Lifetime Education, 2025). Research within a literature review provides evidence for cognitive offloading, suggesting that reliance on AI tool usage reduces critical thinking abilities, while also leading to greater cognitive offloading. For example, Sparrow (et al, 2011), discussed how frequent use of search engines impacted memory, reducing the individual's likelihood of remembering information independently. Another study within Gerlich’s (2025) review found that Zhai (et al, 2024), did a study with students, discovering that students who relied heavily on AI dialogue systems displayed diminished decision-making and critical analysis skills (Gerlich, 2025). 

Consistently, AI responses are exaggerated with rigid structures, providing narrow information—though the responses are vulnerable to being biased due to how information is displayed and captured. This provides users with information that may potentially alter the perception of how the individual views the given information due to the systematic exaggeration and rigid structure of response, which means careful consideration of the AI’s interpretation is optimal (Hanna, et al, 2025).

One study found that artificial personas generated by large language models (such as ChatGPT) were susceptible to cognitive bias through generating human-like text, which can, in turn, impact how humans make decisions and perceive certain subjects. Cognitive biases like the False Consensus Bias, which refers to individuals believing their views are more common than they actually are, can become particularly prevalent through AI use. For example, considering how social media algorithms utilize AI to cater each individual's “for you” page based on their views and interests, this particular cognitive bias is prevalent to consider (Hanna, et al, 2025). 

Another cognitive bias, known as the Anchoring Effect, refers to individuals relying heavily on an initial piece of information when making decisions. Because of this bias, misinformation and misinterpretation can be spread. For example, a study utilizing AI-generated recommended scores on employees overseen by supervisors. Supervisors exposed to high AI-recommended scores rated employee performance much higher than those with low AI- recommended scores. This suggests that AI may amplify or perceive much differently, but still leaves room for more analysis on how AI provides information. Further studies and analyses showed many individuals relied on artificial intelligence to make decisions, which in turn can impact recommendations and influence their final judgements. Using AI as an anchor for decisions and judgements can potentially hinder experiences with the software, such as limitations in data and instead suggests the need for human oversight and additional verification (Carter & Liu, 2025).

Biological Changes

Cognitive functions like decision-making, impulse control, and emotional regulation are negatively impacted through excessive internet use due to grey matter density changes in the brain, specifically in regions like the prefrontal cortex and the anterior cingulate cortex. These areas of the brain reflect memory retrieval demand and goal-relevant control which is focused on decision making and error detection (Shanmugasundaram & Tamilarasu, 2023).

Dergaa (et al., 2024), suggests that the younger generation is especially affected considering their developing brains, leading them to prioritize quick and efficient information instead of promoting critical analysis and deep reflection and comprehension. Instead of cognitive stimulation, they may become overreliant on chatbots, hindering cognitive capabilities and potentially affecting psychological and social impacts. For example, one's mental health may be negatively impacted through erosion of cognitive capabilities leading to reduced self confidence and a sense of helplessness, which may further contribute to cognitive decline. Dergraa (et al, 2024) also mentions that weakened neural pathways, errors in memory retrieval, diminished thinking, decline in ability to apply knowledge, and increased social isolation are some of the other major factors that may occur with AI reliance. 

Cognitive bias can lead to misinformation. However, exposure to misinformation, such as through AI usage, also causes changes in cognition, significantly altering reconstructive processes during memory retrieval. Errors in previously encoded events may be replaced by misleading information, increasing risk for false memories (Lentoor, 2023). Knowing this, it is important to consider information provided by artificial intelligence, understanding the cognitive biases that can occur based on information sought. Further research on how AI and cognitive biases can contribute to changes in the brain may help researchers understand how to prevent or decrease cognitive changes. 

A Glimpse at Social Media and AI Usage

Research has linked the prevalence of AI and social media usage, finding that AI use is positively correlated with social media use. Many social media platforms may rely on AI technology (e.g., recommendations, short descriptions of summaries, tools to ask questions to…etc), which some users may be unaware of. This could cause reliance on AI, potentially unknowingly, and increase the spread of misinformation which can impact cognition (Montag & Elhai, 2025). Pan (et al., (2025) mentions that AI chatbots and GenAI in social media platforms may result in media manipulation, leading to drastic changes in network structures and information. Users should become aware regarding the prevalence of AI chatbots and genAI on social media platforms,  acknowledging  how information is spread and checking sources for accuracy to prevent negative cognitive influence.

Conclusion

Through increased research on cognitive biases, and how artificial intelligence can impact memory, decision-making, critical thinking skills, and other cognitive functions, researchers can better understand how to utilize artificial intelligence without users being susceptible to bias and misinformation. Because artificial intelligence is programmed by humans, there is room for bias and false information within AI responses. Furthermore, given the discussed issues that arise in frequent reliance on artificial intelligence, and how cognitive biases may occur, striving to minimize misinterpretation and deception is ideal, meaning regulation of AI usage should be heavily considered.

References

Hanna Campbell, Samantha Goldman, Patrick M. Markey, Artificial intelligence and human decision making: Exploring similarities in cognitive bias, Computers in Human Behavior: Artificial Humans, Volume 4, 2025, 100138, ISSN 2949-8821, https://doi.org/10.1016/j.chbah.2025.100138.

Lemuria Carter, Dapeng Liu, How was my performance? Exploring the role of

anchoring bias in AI-assisted decision making, International Journal of Information

Management, Volume 82, 2025, 102875, ISSN 0268-4012, https://doi.org/10.1016/j.ijinfomgt.2025.102875.

Da Silva, S., Gupta, R., & Monzani, D. (2023). Editorial: Highlights in psychology:

cognitive bias. Frontiers in psychology, 14, 1242809.

https://doi.org/10.3389/fpsyg.2023.1242809

Dergaa, I., Saad, H. B., Glenn, J. M., Amamou, B., Aissa, M. B., Guelmami, N., Fekih-Romdhane, F., & Chamari, K. (2024). From tools to threats: a reflection on the impact of artificial-intelligence chatbots on cognitive health. Frontiers in Psychology, 15, 1259845. https://doi.org/10.3389/fpsyg.2024.1259845

Duke Learning Innovation & Lifetime Education. (2025, August 13). Does AI harm critical thinking - Duke Center for Teaching and Learning. Duke Center for Teaching and Learning. https://ctl.duke.edu/ai-ethics-learning-toolkit/does-ai-harm-critical-thinking/

Gerlich, M. (2025). AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking. Societies, 15(1), 6. https://doi.org/10.3390/soc15010006

Lentoor A. G. (2023). Cognitive and neural mechanisms underlying false memories: misinformation, distortion, or erroneous configuration?. AIMS neuroscience, 10(3), 255–268. https://doi.org/10.3934/Neuroscience.2023020

Li, X., Zhou, T., Hu, C., & Liu, H. (2025). Availability Bias in AI Recommenders: Impacts on Impulse Buying. Journal of Computer Information Systems, 1–13. https://doi.org/10.1080/08874417.2025.2547173

Liming Pan, Chong-Yang Wang, Fang Zhou, Linyuan Lü, Complexity of social media in the era of generative AI, National Science Review, Volume 12, Issue 1, January 2025, nwae323, https://doi.org/10.1093/nsr/nwae323

Lopez‐Lopez, E., Abels, C. M., Holford, D., Herzog, S. M., & Lewandowsky, S. (2025). Generative artificial intelligence–mediated confirmation bias in health information seeking. Annals of the New York Academy of Sciences, 1550(1), 23–36. https://doi.org/10.1111/nyas.15413

Christian Montag, Jon D. Elhai, The darker side of positive AI attitudes: Investigating associations with (problematic) social media use,mAddictive Behaviors Reports, Volume 22, 2025, 100613, ISSN 2352-8532, https://doi.org/10.1016/j.abrep.2025.100613.

David Paulus, Gerdien de Vries, Marijn Janssen, Bartel Van de Walle, The influence of cognitive bias on crisis decision-making: Experimental evidence on the comparison of bias effects between crisis decision-maker groups, International Journal of Disaster Risk Reduction, Volume 82, 2022, 103379, ISSN 2212-4209, https://doi.org/10.1016/j.ijdrr.2022.103379.

Shanmugasundaram, M., & Tamilarasu, A. (2023). The impact of digital technology, social media, and artificial intelligence on cognitive functions: a review. Frontiers in Cognition, 2. https://doi.org/10.3389/fcogn.2023.1203077

Research guides: Misinformation - Get the Facts: Cognitive Biases. (n.d.). https://guides.lib.uci.edu/Misinfo/bias

Vicente, L., & Matute, H. (2023). Humans inherit artificial intelligence biases. Scientific Reports, 13(1), 15737. https://doi.org/10.1038/s41598-023-42384-8