Cited by 875Ethics and Social Impacts of AIExplainable Artificial Intelligence (XAI)AI and HR Technologies

The DOI matches its Crossref record; 49 of the 59 references it cites were found in a registry or an index; none of the 65 sources its claims cite is checked for support yet.

Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies

Open access

Ferrara, Emilio·2023·Sci

DOI 10.3390/sci6010003

Open this paper in CoChat — free account Open PDF

Abstract

The significant advancements in applying artificial intelligence (AI) to healthcare decision-making, medical diagnosis, and other domains have simultaneously raised concerns about the fairness and bias of AI systems. This is particularly critical in areas like healthcare, employment, criminal justice, credit scoring, and increasingly, in generative AI models (GenAI) that produce synthetic media. Such systems can lead to unfair outcomes and perpetuate existing inequalities, including generative biases that affect the representation of individuals in synthetic data. This survey study offers a succinct, comprehensive overview of fairness and bias in AI, addressing their sources, impacts, and mitigation strategies. We review sources of bias, such as data, algorithm, and human decision biases—highlighting the emergent issue of generative AI bias, where models may reproduce and amplify societal stereotypes. We assess the societal impact of biased AI systems, focusing on perpetuating inequalities and reinforcing harmful stereotypes, especially as generative AI becomes more prevalent in creating content that influences public perception. We explore various proposed mitigation strategies, discuss the ethical considerations of their implementation, and emphasize the need for interdisciplinary collaboration to ensure effectiveness. Through a systematic literature review spanning multiple academic disciplines, we present definitions of AI bias and its different types, including a detailed look at generative AI bias. We discuss the negative impacts of AI bias on individuals and society and provide an overview of current approaches to mitigate AI bias, including data pre-processing, model selection, and post-processing. We emphasize the unique challenges presented by generative AI models and the importance of strategies specifically tailored to address these. Addressing bias in AI requires a holistic approach involving diverse and representative datasets, enhanced transparency and accountability in AI systems, and the exploration of alternative AI paradigms that prioritize fairness and ethical considerations. This survey contributes to the ongoing discussion on developing fair and unbiased AI systems by providing an overview of the sources, impacts, and mitigation strategies related to AI bias, with a particular focus on the emerging field of generative AI.

AI snapshot

Written from the paper’s full text.

TL;DR

This survey reviews sources, impacts, and mitigation strategies for fairness and bias in AI, with particular attention to generative AI bias. It argues that bias arises from data, algorithms, and users, can perpetuate inequality, and requires holistic, interdisciplinary mitigation.

Key findings

Bias in AI is defined as systematic error in decision-making that leads to unfair outcomes; sources include data bias, algorithmic bias, and user bias, with generative bias as an emergent issue in GenAI outputs.
Real-world examples cited include COMPAS being biased against African-American defendants, a healthcare mortality algorithm biased against African-American patients, NIST finding facial recognition less accurate for darker skin tones, and text-to-image models (StableDiffusion, DALL-E, Midjourney) predominantly generating men for CEOs and people of color for criminals or terrorists.
Mitigation approaches reviewed include pre-processing data (oversampling, undersampling, synthetic data, adversarial debiasing), model selection (group fairness, individual fairness, regularization, ensemble methods), and post-processing (e.g., equalized odds); GenAI is said to require holistic strategies, audits, monitoring, and feedback loops.
Fairness types discussed include group fairness (demographic parity, disparate mistreatment, equal opportunity), individual fairness, counterfactual fairness, procedural fairness, and causal fairness; the paper distinguishes fairness as a deliberate goal from bias, which can be unintentional.

Methods

Design: a survey described as a systematic literature review spanning multiple academic disciplines, providing an overview of AI bias sources, impacts, and mitigation strategies.
Data or sample: no primary dataset or sample is analyzed; the paper draws on prior literature and illustrative examples from healthcare, criminal justice, hiring, credit scoring, facial recognition, and generative AI.
Main analysis: narrative synthesis, with tables characterizing types of AI bias, mitigation approaches and challenges, and definitions of fairness; no quantitative meta-analysis or formal search protocol is reported.

Limitations

Stated: lack of diverse and representative training data; difficulty identifying and measuring different types of bias; trade-offs between fairness and accuracy; ethical questions about prioritizing types of bias or groups; and limitations of mitigation methods, including pre-processing being time-consuming or ineffective if data are already biased, model selection lacking consensus on fairness, and post-processing being complex and data-intensive.
Stated: fairness definitions vary and may conflict; group and individual fairness approaches have trade-offs; fairness metrics may not capture intersectionality; and mitigation attempts can have unintended consequences, such as predictive policing efforts potentially increasing racial disparities in arrests.
Evident: as a brief survey, the paper does not report a systematic search strategy, inclusion/exclusion criteria, or quantitative synthesis; its examples are illustrative rather than exhaustive.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/sci6010003

Cites

49 of 50 references verified0 citation errors · 1 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 65 cited sources checked · 0 supported · 0 flagged

In the paper
“Biased algorithms can lead to the underrepresentation of certain groups, such as people of color or those from lower socioeconomic backgrounds, in credit scoring systems, making it harder for them to access loans or mortgages [25].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 3.1. Negative Impacts of Bias in AI on Individuals and Society, Including Discrimination and Perpetuation of Existing Inequalities
What the sources say
- Not checked yetcites [25] Dwork, C., Hardt, M., Pitassi, T., Reingold, O., and Zemel, R. (2012, January 8–10). Fairness through awareness. Proceedings of the 3rd Innovations in Theoretical Computer Science Conference, Cambridge, MA, USA. source
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
In the paper
“The same GenAI models that exhibit gender bias, perhaps unsurprisingly, also portray criminals or terrorists as people of color.”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 3.1. Negative Impacts of Bias in AI on Individuals and Society, Including Discrimination and Perpetuation of Existing Inequalities
What the sources say
- Not checked yetcites [23] Mittelstadt, 2016, The ethics of algorithms: Mapping the debate, Big Data Soc. source
Strength:plain· not yet calibratedCausal:correlational· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
In the paper
“A study by the National Institute of Standards and Technology (NIST) found that facial recognition technology was significantly less accurate for people with darker skin tones, leading to higher rates of false positives [16].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 2.3. Real-World Examples of Bias in AI
What the sources say
- Not checked yetcites [16] Schwartz, R., Vassilev, A., Greene, K., Perine, L., Burt, A., and Hall, P. (2022). Towards a Standard for Identifying and Managing Bias in Artificial Intelligence. source
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
None of the sources cited for this claim have been checked yet.
In the paper
“Researchers and practitioners have proposed various mitigation strategies, such as improving data quality [8] and designing explicitly fair algorithms [9][10][11].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 1. Introduction
What the sources say
- Not checked yetcites [8] Asan, 2020, Artificial intelligence and human trust in healthcare: Focus on clinicians, J. Med. Internet Res. source
- Not checked yetcites [9][10][11]
Strength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
In the paper
“Furthermore, when prompted to generate images of criminals or terrorists, the models' output overwhelmingly more people of color.”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 2.3. Real-World Examples of Bias in AI
What the sources say
- Not checked yetcites [23] Mittelstadt, 2016, The ethics of algorithms: Mapping the debate, Big Data Soc. source
Strength:boosted· not yet calibratedCausal:correlational· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
In the paper
“A striking instance of GenAI bias was reported, where text-to-image models like StableDiffusion, OpenAI's DALL-E, and Midjourney exhibited racial and stereotypical biases in their outputs [23].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 2.3. Real-World Examples of Bias in AI
What the sources say
- Not checked yetcites [23] Mittelstadt, 2016, The ethics of algorithms: Mapping the debate, Big Data Soc. source
Strength:plain· not yet calibratedCausal:correlational· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
None of the sources cited for this claim have been checked yet.
In the paper
“Multiple studies have identified biases against certain groups in AI systems, such as the facial recognition systems studied by Buolamwini and Gebru [1] and hiring algorithms examined by Dastin [2].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 1. Introduction
What the sources say
- Not checked yetcites [1] Buolamwini, J., and Gebru, T. (2018, January 23–24). Gender shades: Intersectional accuracy disparities in commercial gender classification. Proceedings of the 1st Conference on Fairness, Accountability and Transparency, New York, NY, USA.
- Not checked yetcites [2] Dastin, J. (2018). Ethics of Data and Analytics, Auerbach Publications.
Load-bearingStrength:plain· not yet calibratedCausal:correlational· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes

Show all 32 claims

It may state two separate claims as one. 26
Probably not a substantial claim. 24
The quote only partly states this claim. 1

The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Additionally, there may be privacy concerns when collecting certain types of data, such as medical records or financial information.”
- Also stated in: § 2. Unintended consequences on the distribution of outcomes · plain
  “These challenges can limit the effectiveness of dataset augmentation as a mitigation approach.”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 2. Unintended consequences on the distribution of outcomes
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
The quote only partly states this claim.
In the paper
“These biases can perpetuate systemic discrimination and inequality, with detrimental effects on individuals and communities in areas like hiring, lending, and criminal justice [3][4][5][6][7].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 1. Introduction
What the sources say
- Not checked yetcites [3][4][5][6][7]
Load-bearingStrength:plain· not yet calibratedCausal:conditional· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“In the realm of generative AI, addressing bias is even more challenging as it requires a holistic strategy [14,21,22].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 4.1. Overview of Current Approaches to Mitigate Bias in AI, Including Pre-Processing Data, Model Selection, and Post-Processing Decisions
What the sources say
- Not checked yetcites [14] Ferrara, E. (2023). Should ChatGPT be Biased? Challenges and Risks of Bias in Large Language Models. First Monday, 28. source
- Not checked yetcites [21] Ferrara, E. (2023). GenAI against humanity: Nefarious applications of generative artificial intelligence and large language models. arXiv. source
- Not checked yetcites [22] Ferrara, E. (2023). The butterfly effect in artificial intelligence systems: Implications for AI bias and fairness. arXiv. source
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✗ Checkable: fails✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. It may be too vague to check against evidence. The paper’s conclusions may not depend on it.
In the paper
“The importance of fairness and bias in AI is widely recognized by researchers, policymakers, and the academic community [1,[12][13][14][15][16].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 1. Introduction
What the sources say
- Not checked yetcites [1,[12][13][14][15][16]
Strength:boosted· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“For instance, facial recognition algorithms trained on data primarily consisting of men can struggle to recognize female faces accurately, perpetuating gender bias in security systems [1].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 3.1. Negative Impacts of Bias in AI on Individuals and Society, Including Discrimination and Perpetuation of Existing Inequalities
What the sources say
- Not checked yetcites [1] Buolamwini, J., and Gebru, T. (2018, January 23–24). Gender shades: Intersectional accuracy disparities in commercial gender classification. Proceedings of the 1st Conference on Fairness, Accountability and Transparency, New York, NY, USA.
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“Additionally, model selection methods may be limited by the lack of consensus on what constitutes fairness, and post-processing methods can be complex and require large amounts of additional data [32].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 4.1. Overview of Current Approaches to Mitigate Bias in AI, Including Pre-Processing Data, Model Selection, and Post-Processing Decisions
What the sources say
- Not checked yetcites [32] Barocas, 2016, Big data’s disparate impact, Calif. Law Rev.
Strength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“A study by ProPublica found that the system was biased against African-American defendants, as they were more likely to be labeled as high-risk even if they had no prior convictions.”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 2.3. Real-World Examples of Bias in AI
What the sources say
- Not checked yetcites ProPublica
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“For example, biased algorithms used in the criminal justice system can lead to unfair treatment of certain groups, particularly people of color, who are more likely to be wrongly convicted or receive harsher sentences [1].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 3.1. Negative Impacts of Bias in AI on Individuals and Society, Including Discrimination and Perpetuation of Existing Inequalities
What the sources say
- Not checked yetcites [1] Buolamwini, J., and Gebru, T. (2018, January 23–24). Gender shades: Intersectional accuracy disparities in commercial gender classification. Proceedings of the 1st Conference on Fairness, Accountability and Transparency, New York, NY, USA.
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“Different people and groups may have different definitions of fairness, and these definitions may change over time [39].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 6.2. Discussion of the Limitations and Challenges of These Approaches
What the sources say
- Not checked yetcites [39] Ezzeldin, Y.H., Yan, S., He, C., Ferrara, E., and Avestimehr, S. (2023, January 7–14). Fairfed: Enabling group fairness in federated learning. Proceedings of the AAAI 2023—37th AAAI Conference on Artificial Intelligence, Washington, DC, USA.
Strength:hedged· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“This can be performed through techniques such as regularization, which penalizes models for making discriminatory predictions, or through ensemble methods, which combine multiple models to reduce bias [25].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 4.1. Overview of Current Approaches to Mitigate Bias in AI, Including Pre-Processing Data, Model Selection, and Post-Processing Decisions
What the sources say
- Not checked yetcites [25] Dwork, C., Hardt, M., Pitassi, T., Reingold, O., and Zemel, R. (2012, January 8–10). Fairness through awareness. Proceedings of the 3rd Innovations in Theoretical Computer Science Conference, Cambridge, MA, USA. source
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Other approaches include transparency, accountability, and explainability.”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 6.1. Overview of Current Approaches to Ensure Fairness in AI, including Group Fairness and Individual Fairness
What the sources say
- Not checked yetcites [26] Ananny, 2018, Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability, New Media Soc. source
- Not checked yetcites [38] Donovan, J., Caplan, R., Matthews, J., and Hanson, L. (2018). Algorithmic Accountability: A Primer, Data & Society.
Strength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“Additionally, different types of fairness may conflict with each other, and trade-offs may need to be made to achieve fairness in specific contexts [26].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.1. Definition of Fairness in AI and Its Different Types
What the sources say
- Not checked yetcites [26] Ananny, 2018, Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability, New Media Soc. source
Strength:hedged· not yet calibratedCausal:conditional· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Additionally, it may be difficult to determine which types of fairness are most appropriate for a given context and how to balance them appropriately [4][5][6].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 6.2. Discussion of the Limitations and Challenges of These Approaches
What the sources say
- Not checked yetcites [4] Kleinberg, 2018, Human decisions and machine predictions, Q. J. Econ.
- Not checked yetcites [5] Kleinberg, 2018, Discrimination in the Age of Algorithms, J. Leg. Anal. source
- Not checked yetcites [6] Kleinberg, 2018, Algorithmic fairness, AEA Pap. Proc. source
Strength:hedged· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“To address this issue, the Northpointe COMPAS was modified to include a “race-neutral” version of the algorithm that achieved similar accuracy while reducing racial bias [19].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.3. Real-World Examples of Fairness in AI
What the sources say
- Not checked yetcites [19] Angwin, J., Larson, J., Mattu, S., and Kirchner, L. (2016). Ethics of Data and Analytics, Auerbach Publications.
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“To address this issue, researchers have proposed using techniques such as subgroup analysis to identify and address bias in the data used to train AI models [11].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.3. Real-World Examples of Fairness in AI
What the sources say
- Not checked yetcites [11] Yan, S., Kao, H.T., and Ferrara, E. (July, January 26). Fair class balancing: Enhancing model fairness without observing sensitive attributes. Proceedings of the 29th ACM International Conference on Information & Knowledge Management, Online.
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“In other words, bias can be viewed as a technical issue, while fairness is a social and ethical issue [26].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.2. Comparison of Fairness and Bias in AI
What the sources say
- Not checked yetcites [26] Ananny, 2018, Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability, New Media Soc. source
Strength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
- Rests on an assumption: Bias can be unintentional, while fairness is inherently a deliberate and intentional goal.
In the paper
“There are several different types of fairness that have been proposed in the literature, including group fairness, individual fairness, and counterfactual fairness [30].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.1. Definition of Fairness in AI and Its Different Types
What the sources say
- Not checked yetcites [30] Zafar, M.B., Valera, I., Gomez Rodriguez, M., and Gummadi, K.P. (2017, January 3–7). Fairness beyond disparate treatment & disparate impact: Learning classification without disparate mistreatment. Proceedings of the 26th International Conference on World Wide Web, Perth, Australia. source
Strength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“One example is the COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) system, which is used to predict the likelihood of recidivism in criminal defendants.”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.3. Real-World Examples of Fairness in AI
What the sources say
- Not checked yetcites [19] Angwin, J., Larson, J., Mattu, S., and Kirchner, L. (2016). Ethics of Data and Analytics, Auerbach Publications.
Strength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. The paper’s conclusions may not depend on it.
None of the sources cited for this claim have been checked yet.
In the paper
“Diversifying the teams involved in AI development and evaluation is equally crucial, as it brings a multiplicity of perspectives that can better identify and correct for biases [55][56][57].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 7. Conclusions
What the sources say
- Not checked yetcites [55] Chauhan, 2022, The Role of Data and Artificial Intelligence in Driving Diversity, Equity, and Inclusion, Computer source
- Not checked yetcites [56] Holstein, K., Wortman Vaughan, J., Daumé, H., Dudik, M., and Wallach, H. (2019, January 4–9). Improving fairness in machine learning systems: What do industry practitioners need?. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems, Glasgow, UK. source
- Not checked yetcites [57] Stathoulopoulos, K., Mateos-Garcia, J.C., and Owen, H. (2023, December 15). Gender Diversity in AI Research. Available online: https://www.nesta.org.uk/report/gender-diversity-ai/.
Load-bearingStrength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✗ Checkable: fails✓ Load-bearing: passes✓ Based: passes
Probably not a substantial claim. It may be too vague to check against evidence.
In the paper
“Another difference is that bias can be either positive or negative, whereas fairness is only concerned with negative bias or discrimination [36].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.2. Comparison of Fairness and Bias in AI
What the sources say
- Not checked yetcites [36] 2017, Measuring discrimination in algorithmic decision making, Data Min. Knowl. Discov. source
Strength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✗ Load-bearing: fails✓ Based: passes
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“We also considered the ethical implications of AI in preserving privacy and the necessity for transparency, oversight, and continuous evaluation of AI systems [52][53][54].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 7. Conclusions
What the sources say
- Not checked yetcites [52] Lipton, 2018, The mythos of model interpretability: In machine learning, the concept of interpretability is both important and slippery, Queue source
- Not checked yetcites [53] Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman, L., Hutchinson, B., Spitzer, E., Raji, I.D., and Gebru, T. (2019, January 29–31). Model cards for model reporting. Proceedings of the Conference on Fairness, Accountability, and Transparency, Atlanta, GA, USA. source
- Not checked yetcites [54] Raji, I.D., and Buolamwini, J. (2019, January 29–31). Actionable auditing: Investigating the impact of publicly naming biased performance results of commercial AI products. Proceedings of the Conference on Fairness, Accountability, and Transparency, Atlanta, GA, USA. source
Load-bearingStrength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes
Probably not a substantial claim. It may state two separate claims as one.
In the paper
“Research has shown that the system was biased against African-American defendants, as it was more likely to falsely predict that they would re-offend compared to white defendants [19].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.3. Real-World Examples of Fairness in AI
What the sources say
- Not checked yetcites [19] Angwin, J., Larson, J., Mattu, S., and Kirchner, L. (2016). Ethics of Data and Analytics, Auerbach Publications.
Strength:plain· not yet calibratedCausal:direct· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes? Load-bearing: unclear✓ Based: passes
It may state two separate claims as one.
In the paper
“However, it is important to recognize that bias is not the only factor that can lead to unfairness, and achieving fairness may require additional efforts beyond bias mitigation [27].”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 5.2. Comparison of Fairness and Bias in AI
What the sources say
- Not checked yetcites [27] Corbett-Davies, S., Pierson, E., Feller, A., Goel, S., and Huq, A. (2017, January 13–17). Algorithmic decision making and the cost of fairness. Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, NS, Canada. source
Load-bearingStrength:plain· not yet calibratedCausal:conditional· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes
It may state two separate claims as one.
- Rests on an assumption: Addressing bias in training data or algorithms can help reduce the likelihood of unfair outcomes.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“While these approaches have shown promising results in promoting fairness in AI, they are not without limitations and challenges.”
§ Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies > 6.2. Discussion of the Limitations and Challenges of These Approaches
Load-bearingStrength:plain· not yet calibrated
Tests✓ Asserted: passes✓ Checkable: passes✓ Load-bearing: passes✓ Based: passes
It may state two separate claims as one.
- Rests on an assumption: Many current approaches to ensuring fairness in AI rely on statistical methods and assumptions that may not accurately capture the complexity of human behavior and decision-making.

41 of 41

The paper relies on this without support.
The use of AI systems is growing.
“The growing use of AI systems has intensified discussions around fairness and bias in artificial intelligence as potential biases and discrimination become more apparent.”
The paper relies on this without support.
Generative AI is becoming more prevalent in creating content that influences public perception.
“especially as generative AI becomes more prevalent in creating content that influences public perception.”
The paper relies on this without support.
Bias can be unintentional, while fairness is inherently a deliberate and intentional goal.
“One key difference between fairness and bias is that while bias can be unintentional, fairness is inherently a deliberate and intentional goal.”
The paper relies on this without support.
Addressing bias in training data or algorithms can help reduce the likelihood of unfair outcomes.
“For example, addressing bias in training data or algorithms can help reduce the likelihood of unfair outcomes.”
The paper relies on this without support.
Group fairness can be achieved through various techniques such as re-sampling, pre-processing, or post-processing of the data used to train the AI model.
“This can be achieved through various techniques such as re-sampling, pre-processing, or postprocessing of the data used to train the AI model.”
The paper relies on this without support.
Individual fairness can be achieved through techniques such as counterfactual fairness or causal fairness.
“Individual fairness can be achieved through techniques such as counterfactual fairness or causal fairness.”
The paper relies on this without support.
Many current approaches to ensuring fairness in AI rely on statistical methods and assumptions that may not accurately capture the complexity of human behavior and decision-making.
“Furthermore, many of the current approaches to ensuring fairness in AI rely on statistical methods and assumptions that may not accurately capture the complexity of human behavior and decision-making.”
The paper relies on this without support.
Bias can arise due to various factors, such as biased data or algorithmic design.
“Bias can arise due to various factors, such as biased data or algorithmic design, but fairness requires a conscious effort to ensure that the algorithm does not discriminate against any group or individual.”
The paper relies on this without support.
Fairness requires a conscious effort to ensure that the algorithm does not discriminate against any group or individual.
“Bias can arise due to various factors, such as biased data or algorithmic design, but fairness requires a conscious effort to ensure that the algorithm does not discriminate against any group or individual.”
The paper relies on this without support.
If an AI model is trained on data that are biased toward a particular group, re-sampling techniques can be used to create a balanced dataset where each group is represented equally.
“For example, if an AI model is trained on data that are biased toward a particular group, re-sampling techniques can be used to create a balanced dataset where each group is represented equally.”
The paper relies on this without support.
Pre-processing or post-processing can be used to adjust the output of the AI model to ensure that it does not unfairly disadvantage any group.
“Other techniques, such as pre-processing or post-processing, can be used to adjust the output of the AI model to ensure that it does not unfairly disadvantage any group.”

11 of 11

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

Fairness both excludes and includes favoritism
Minor
The paper says fairness is concerned only with negative bias, but elsewhere defines fairness as the absence of discrimination or favoritism. It also defines positive bias as systematically favoring a group, so the stated scope of fairness conflicts with its definition.
Model training classified as pre-processing
Minor
The paper defines pre-processing as addressing bias before the model is trained, then lists a technique involving training the model as an example of pre-processing.

2 of 2

Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory, and Antiracist Politics
Crenshaw, K. (2018) · Living with Contradictions · doi:10.4324/9780429499142-5 · Cited by 7,232
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137 · cited 1×
“A study conducted by Obermeyer et al. [20] found that the system was more likely to assign higher-risk scores to African-American patients, even when other factors, such as age and health status, were the same.”
A Survey on Bias and Fairness in Machine Learning
Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., & Galstyan, A. (2022) · ACM Computing Surveys · doi:10.1145/3457607 · Cited by 5,812
The Mythos of Model Interpretability: In machine learning, the concept of interpretability is both important and slippery.
Lipton, Z. C. (2018) · Queue · doi:10.1145/3236386.3241340 · Cited by 5,427
The ethics of algorithms: Mapping the debate
Mittelstadt, B. D., Allo, P., Taddeo, M., Wachter, S., & Floridi, L. (2016) · Big Data & Society · doi:10.1177/2053951716679679 · Cited by 4,567 · cited 3×
“When generative AI (GenAI) models are prompted to create images of CEOs, they tend to reinforce stereotypes by depicting CEOs predominantly as men [23].”
BackgroundMethodInfluential
Untitled work
(2018) · Cited by 3,510
Fairness through awareness
Dwork, C., Hardt, M., Pitassi, T., Reingold, O., & Zemel, R. (2012) · Proceedings of the 3rd Innovations in Theoretical Computer Science Conference · doi:10.1145/2090236.2090255 · Cited by 3,501 · cited 4×
“Similarly, an image generation model trained on datasets with limited diversity in human portraits may struggle to accurately represent a broad range of ethnicities.”
BackgroundMethodInfluential
Big Data's Disparate Impact
Barocas, S., & Selbst, A. D. (2014) · doi:10.2139/ssrn.2477899 · Cited by 3,048 · cited 2×
“Can be complex and require large amounts of additional data [32].”
Background
Semantics derived automatically from language corpora contain human-like biases
Caliskan, A., Bryson, J. J., & Narayanan, A. (2017) · Science · doi:10.1126/science.aal4230 · Cited by 2,887 · cited 1×
“This survey study delves into the complex and multifaceted issues surrounding fairness and bias in AI, covering the sources of bias, their impacts, and proposed mitigation strategies.”
Background
Algorithms of Oppression: How Search Engines Reinforce Racism
Noble, S. U. (2020) · doi:10.2307/j.ctt1pwt9w5 · Cited by 2,709

Flagged first, then by citations

Show all 59 references

9 of the 59 references couldn’t be matched to a known paper.

Research works printed but not linked: 3

Eubanks, V. (2018). Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor, St. Martin’s Press.
Corbett-Davies, S., and Goel, S. (2018). The measure and mismeasure of fairness: A critical review of fair machine learning. arXiv.
Crenshaw, K. (1989). Feminist Legal Theories, Routledge.

Sorted by citations

Items 301–400 of 842

AI-Driven Innovations in Personalized Radiopharmaceutical Therapy
Karthikeyan, B., Dhanaselvam, P. S., Kavitha, K., Nagarani, N., Karthikeyan, P., & Kumar, G. P. (2025) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-1275-0.ch011 · Cited by 4
Algorithmic discrimination and explainable artificial intelligence in the judiciary: a case study of the Constitutional Court of Colombia
Tapias, B. A. H., Guzmán, D. H., Zambam, N. J., Turatti, L., Rodríguez, F. A., Fröhlich, S., Staffen, M. R., Muñoz, P. M. C., Reyes, A. G., Soto, G., et al. (2025) · Procedia Computer Science · doi:10.1016/j.procs.2025.03.164 · Cited by 10
Machine Learning and Artificial Intelligence in Suicide Prevention: A Bibliometric Analysis of Emerging Trends and Implications for Nursing
Yıldız, E. (2025) · Issues in Mental Health Nursing · doi:10.1080/01612840.2025.2505904 · Cited by 6
“In particular, issues of bias in systems have become more complex with the use of generative models (Ferrara, 2023).”
Saudi Arabia's Vision 2030: Leveraging Generative Artificial Intelligence to Enhance Software Engineering
Albaroudi, E., Mansouri, T., Hatamleh, M., et al. (2025) · doi:10.1109/wids-psu64963.2025.00024 · Cited by 5
“Opaque AI systems may undermine public trust in technology, causing backlash from the citizens like legal systems using AI-generated sentencing guidelines without proper reasoning [31], [34].”
Raising awareness of potential biases in medical machine learning: Experience from a Datathon
Hochheiser, H., Klug, J., Mathie, T., Pollard, T., Raffa, J. D., Ballard, S. L., Conrad, E. A., Edakalavan, S., Joseph, A. M., Alnomasy, N., et al. (2025) · PLOS Digital Health · doi:10.1371/journal.pdig.0000932 · Cited by 4
A Statistical Analysis to Investigate the Factors Affecting Generative AI Use in Education and Its Impacts on Social Sustainability Using SPSS
Binhammad, M. H. Y., & Shaalan, K. F. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-65996-6_15 · Cited by 4
Addressing the Challenges of Auditing and Testing for AI Bias: A Comparative Analysis of Regulatory FrameworksPreprint
Krause, D. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5050631 · Cited by 4
Academic Integrity and Students’ Ethical Use of ChatGPT in Higher Education
Shishakly, R., & Nachouki, M. (2026) · Journal of Information Technology Education: Research · doi:10.28945/5730 · Cited by 15
“…academic integrity, trust, plagiarism, algorithmic bias, perceived usefulness, and digital literacy (Bissessar, 2025; Chan & Hu, 2023; Chaudhry et al., 2022; Ferrara, 2024; Kotsis, 2024; Ma’amor et al., 2024; Pacheco - Mendoza et al., 2023; Prashar et al., 2024; Vieriu & Petrea, 2025; Zhai, 2022).”
Developer Perspectives on Licensing and Copyright Issues Arising from Generative AI for Software Development
Stalnaker, T., Wintersgill, N., Chaparro, O., Heymann, L. A., Penta, M. D., Germán, D. M., & Poshyvanyk, D. (2025) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3743133 · Cited by 17
Key Drivers of Artificial Intelligence on Learner Engagement and Experience
Iyer, S. S. (2024) · Advances in educational technologies and instructional design book series · doi:10.4018/979-8-3693-7220-3.ch004 · Cited by 3
Ethical Considerations and Bias in AI Applications in Healthcare
Ali, A. (2025) · Information systems engineering and management · doi:10.1007/978-3-031-84404-1_7 · Cited by 4
Artificial Intelligence in Fire Safety: A Critical Perspective on Policy, Stakeholders and Emerging Technologies in India
Kumar, R., Kaur, A., Dangi, H. K., Kumari, P., & Kumar, N. (2025) · FIIB Business Review · doi:10.1177/23197145251342714 · Cited by 3
“These issues are to be further investigated so that these systems are unbiased, mainly in the context of economically disadvantaged areas (Ferrara, 2023).”
Influential
XAI4RE – Using Explainable AI for Responsible and Ethical AI
Tal, A. S., & Sheidin, J. (2025) · ACM Conference on User Modeling, Adaptation and Personalization (UMAP) · doi:10.1145/3708319.3733707 · Cited by 3
“For end-users and decision-makers, XAI can improve their understanding, trust, and task performance [39, 41] and facilitate human-AI collaboration in the decision-making process [12, 13, 36].”
Reconsidering Requirements Engineering: Human–AI Collaboration in AI-Native Software Development
Abbasi, M. A., Ihantola, P., Mikkonen, T., & Mäkitalo, N. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04190-6_11 · Cited by 6
“What new challenges AI has introduced: AI tools and techniques can enhance prioritization and stakeholder involvement, but it also raises issues such as bias, transparency, and ethical considerations [14,60,42,61].”
Influential
AI-Augmented Prevention Science Needs Community-Engaged Prevention Science: a Framework for Greater Accountability
Rice, E. R. W., & Young, L. E. (2025) · Prevention Science · doi:10.1007/s11121-025-01842-7 · Cited by 3
“Bias in AI can be defined as systematic errors in algorithmic decision-making processes that result in unfair outcomes (Ferrara, 2023).”
FAIR-MED: Bias Detection and Fairness Evaluation in Healthcare Focused XAI
Bahamazava, K., & O’Reilly, R. (2025) · Communications in computer and information science · doi:10.1007/978-3-032-08317-3_18 · Cited by 3
Toward emotionally intelligent interfaces: An HCI approach to cognitive and learning support
Kumar, P., Kurnianto, A. A., & Sahai, N. (2025) · Elsevier eBooks · doi:10.1016/b978-0-443-41660-6.00015-6 · Cited by 3
Stakeholders’ perceived benefits and concerns regarding artificial intelligence in the care of older adults
Chae, K., Massare, J., Ashida, S., Cudjoe, T., Abadir, P., Arbaje, A. I., Unberath, M., Phan, P. H. C., & Schoenborn, N. (2025) · Journal of The American Geriatrics Society · doi:10.1111/jgs.70228 · Cited by 3
“Concern for AI applications to introduce bias has been previously described, which mostly focused on bias across racial, ethnic, or cultural groups [28, 29].”
Creativity in the age of generative AI
Yasuda, A., & Maruyama, Y. (2025) · AI and Ethics · doi:10.1007/s43681-025-00848-9 · Cited by 7
“For example, generative AI can influ - ence not only the content but also the tone and context of the generated text or images [53].”
EEG-based Alzheimer’s disease detection: Past, present, and future
Alidadi, M., Hamavar, R., Abbasi, E., Nabavifard, S., Mohebbi, M., & Asl, B. M. (2025) · Measurement · doi:10.1016/j.measurement.2025.120096 · Cited by 4
Generative Artificial Intelligence Use in Transactional Analysis Psychotherapy: Transference With the Artificial Third
Haworth, M. (2026) · Transactional Analysis Journal · doi:10.1080/03621537.2026.2595388 · Cited by 4
“…available and used to train them with. this has resulted in many models being Anglocentric—for example, giving more accurate and complete information if asked in english rather than in Arabic (Sallam et al., 2024) and displaying biases such as sexism and racism (Ferrara, 2023; Zhou et al., 2024).”
What is a fact in the age of generative AI? Fact-checking as an epistemological lens
Dierickx, L., Opdahl, A. L., Bjerknes, F., & Lindén, C. (2026) · Information Communication & Society · doi:10.1080/1369118x.2026.2630697 · Cited by 8
“Biases may be representational, technical, or social in nature (e.g., Dwivedi et al., 2023; Ferrara, 2024; Gallegos et al., 2024; Ruggieri et al., 2024), and cannot be captured solely through output evaluation.”
Addressing Bias and Ensuring Fairness in AI Systems for Healthcare
Zarnegar, A. (2025) · Health Informatics and Knowledge Management Conference (HIKM) · doi:10.1145/3774816.3774817 · Cited by 3
Nanobots in medicine and beyond: From targeted drug delivery to intelligent environmental remediation
Bhushan, R., Sonker, P., Singh, P., et al. (2026) · Next Materials · doi:10.1016/j.nxmate.2026.101791 · Cited by 3
How do large-language models respond to moral dilemmas? Insights from the defining issues test
Nabizadeh, D., Walker, D. I., Han, H., McNeill, L., Wind, S. A., & Scofield, J. M. (2026) · AI and Ethics · doi:10.1007/s43681-026-01104-4 · Cited by 5
Explainable AI in Medicine: A Comprehensive Narrative Review of Methods, Applications, and Future Directions
Wang, H. J., Shi, G., Liu, X., Zhang, J., Chen, S., Shi, J., Li, Y., Yue, G., & Wu, Y. (2026) · Expert Systems · doi:10.1111/exsy.70259 · Cited by 5
“This calls for new techniques that show where and why performance diverges between groups, such as subgroup conditioned attribution, counterfactual reasoning, or group specific concept analyses, while avoiding the disclosure of identifiable characteristics (Ferrara 2024).”
An Exploratory Analysis of using Chatbots in Academia
Al-harbi, N. K., & Al-Shargabi, A. A. (2023) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2023.0141212 · Cited by 3
“Machine learning models, a type of artificial intelligence system can learn, and replicate bias patterns found in the training data, which leads to unfair and discriminatory results [25].”
Background
Defenses Against Artificial Intelligence Attacks
Tsikerdekis, M., Zeadally, S., & Katib, I. A. (2024) · Computer · doi:10.1109/mc.2024.3420782 · Cited by 3
“10 This data sensitivity vulnerability arises from the model’s susceptibility to the addition of skewed information to its training data during the learning process.”
Background
Appropriation of Artificial Intelligence in Broadcast Media Production in Kenya: Opportunities and Concerns
Muya, P. (2024) · Journal of the Kenya National Commission for UNESCO · doi:10.62049/jkncu.v5i1.202 · Cited by 5
Enhancements for Developing a Comprehensive AI Fairness Assessment Standard
Agarwal, A., Kumar, M., & Nene, M. J. (2025) · doi:10.1109/comsnets63942.2025.10885551 · Cited by 4
Generative Artificial Intelligence
Babu, B. S., Umakanth, M., & Bhanumathi, P. (2025) · IGI Global eBooks · doi:10.4018/979-8-3693-7398-9.ch020 · Cited by 3
Securing Digital Transformation
Karakaya, T. (2025) · IGI Global eBooks · doi:10.4018/979-8-3693-6417-8.ch006 · Cited by 3
Exploring the Creative Potential of AI in Filmmaking
Pan, Z., Benini, S., Grierson, M., Savardi, M., & Smith, T. J. (2025) · Creativity and Cognition · doi:10.1145/3698061.3728393 · Cited by 3
AI biases as asymmetries: a review to guide practice
Waters, G., & Honenberger, P. (2025) · Frontiers in Big Data · doi:10.3389/fdata.2025.1532397 · Cited by 7
“A recent systematic review deﬁnes “bias” as “systematic error in decision-making processes that results in unfair outcomes” (Ferraro, 2024).”
Generative artificial intelligence and the future of financial forecasting: Evidence from large language models
Akpan, M. N. U. (2025) · Risk Governance and Control Financial Markets & Institutions · doi:10.22495/rgcv15i4p13 · Cited by 3
Exploring Ethical Dimensions of AI Assistants and Chatbots
Porna, S. B., Ahmad, M., Vallejo, R. G., Shahzadi, I., & Rahman, M. A. (2024) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3693-9173-0.ch011 · Cited by 4
Beyond Accuracy: Rethinking Outlier Detection in Asylum Data
Kaltenhauser, K., Slaats, T., Muller, M. J., & Møller, N. H. (2025) · ACM Journal on Responsible Computing · doi:10.1145/3718986 · Cited by 4
Legal Implications of AI-Driven OSINT: Insider Threats and Data Leaks in Egypt and the European UnionPreprint
Kamal, M. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4672751 · Cited by 3
Assessing Trustworthy Ai: Technical and Legal Perspectives of Fairness in AiPreprint
Kattnig, M., Angerschmid, A., Reichel, T., et al. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4705900 · Cited by 3
AI in Public Governance: A Reality Check on Scholarly Insights and Policy PracticePreprint
Helmholz, P., Nolte, M., & Schmitt, M. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4787049 · Cited by 4
XtremeLLMs: Towards Extremely Large Language ModelsPreprint
Mienye, I. D., Swart, T. G., & Obaido, G. (2024) · Preprints.org · doi:10.20944/preprints202408.1483.v1 · Cited by 3
Experimental economics for machine learning—a methodological contribution on lie detection
Bershadskyy, D., Dinges, L., Fiedler, M.-A., Al-Hamadi, A. K., Ostermaier, N., & Weimann, J. (2024) · PLoS ONE · doi:10.1371/journal.pone.0314806 · Cited by 2
A Review on Conversational AI as a Tool in Academic Writing
Demiröz, H., & Tıkız‐Ertürk, G. (2025) · Eskiyeni (Online)/Eskiyeni · doi:10.37697/eskiyeni.1565854 · Cited by 6
Human-Technology Interaction
Thakur, S., Singh, H., Mishra, N., Tripathy, R., Nayak, R. K., & Thakur, A. (2025) · Advances in educational technologies and instructional design book series · doi:10.4018/979-8-3373-2262-9.ch013 · Cited by 2
SINIR GÜVENLİĞİNDE YAPAY ZEKÂ ENTEGRASYONU
Uzunparmak, M., & İrdem, İ. (2025) · Güvenlik Bilimleri Dergisi · doi:10.28956/gbd.1569382 · Cited by 2
A STUDY ON HYBRID RECOMMENDER SYSTEMS FOR EFFECTIVE TARGETED MARKETING IN E-COMMERCE PLATFORMS
Chakraborty, S. (2025) · International Journal of Advances in Business and Management Research · doi:10.62674/ijabmr.2025.v2i04.006 · Cited by 2
Tech Meets Tradition: A Nigerian Perspective on Aligning AI Development with African Religious Values
Aluko, O. (2025) · doi:10.64553/dnao2503 · Cited by 1
GenAI Use Behavior and Post‐Failure Perceptions Among People With Functional Disabilities: A Multimethod Study
Bagnato, G., Moreno, F. R., Campayo-Sánchez, F., & Ali, A. A. (2025) · Psychology and Marketing · doi:10.1002/mar.70028 · Cited by 4
“These coping strategies help them manage frustration, protect their emotional well ‐ being, and maintain engagement with the technology despite its flaws (Shinohara and Wobbrock 2011; Ferrara 2023).”
A Review of Librarians as Digital Literacy Drivers in an Age of Generative AI Surrounded by Ethical Conundrums
Onwubiko, E. C. (2025) · Frontiers in Educational Practice and Research · doi:10.55578/fepr.2507.006 · Cited by 2
Revolutionizing Education with AI: Ethical Considerations in K-12 Settings
Gökçe, A. T. (2025) · Education and human development · doi:10.5772/intechopen.1012230 · Cited by 2
Ethical AI and Sustainable Digital Governance: An Interdisciplinary Framework
Sajjanapu, M. (2025) · European Modern Studies Journal · doi:10.59573/emsj.9(4).2025.76 · Cited by 2
Comprehensive Bias Mitigation in AI: Evaluating Pre-Processing, In-Processing, and Post-Processing Techniques for Fair Decision-Making
Skaiky, A. A., Ali, H. M. S., Mohammed, A., et al. (2025) · IEEE International Conference on Computing and Machine Intelligence (ICMI) · doi:10.1109/icmi65310.2025.11141086 · Cited by 2
“Ferrara [18] gave a complete overview of fairness and bias in AI, and especially in sensitive fields around healthcare, employment, and criminal justice.”
Biases in AI-Supported Industry 4.0 Research: A Systematic Review, Taxonomy, and Mitigation Strategies
Arévalo-Royo, J., Montalvo, F. J. F., Latorre‐Biel, J., Macías, E. J., Martínez, E., & Blanco‐Fernández, J. (2025) · Applied Sciences · doi:10.3390/app152010913 · Cited by 3
The impact of AI adoption on organizational performance: ethical and technological factors in Iran’s startup ecosystem
Salamzadeh, A., Ashkani, M., & Asgharifar, M. H. (2025) · Strategy and Leadership · doi:10.1108/sl-04-2025-0091 · Cited by 6
Participatory modeling in the AI era
Kolagani, N., Glynn, P. D., Voinov, A. A., Quinn, N. W. T., HELGESON, J. F., & Dyckman, C. S. (2025) · Environmental Modelling & Software · doi:10.1016/j.envsoft.2025.106762 · Cited by 2
Trustworthy requirements for foundation models—A comprehensive survey and roadmap
Song, P., Ojo, A., & Curry, E. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.113111 · Cited by 5
A theoretical model of AI bias mitigation: incentives, regulation and equilibrium
Bera, S., & Bera, I. (2025) · Journal of Information Communication and Ethics in Society · doi:10.1108/jices-06-2025-0138 · Cited by 2
ChatGPT-4.0 as a Tool for Automated Review of Ethics and Transparency in Biomedical Literature
Doskaliuk, B., Seiil, B., & Qumar, A. B. (2025) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e24 · Cited by 3
AI Voice Methodology: Using Generative AI in Qualitative Social Research
Hughes, J., Homan, L., O’Reilly, M., & Hughes, K. (2025) · Qualitative Inquiry · doi:10.1177/10778004251401842 · Cited by 2
Re-orienting science education in the generative AI era: From inquiry to practice
Zhong, B., Xiaohong, L., & Liu, X. (2026) · Innovations in Education and Teaching International · doi:10.1080/14703297.2026.2614480 · Cited by 2
Redefining Participation: A Generative Approach to Elevating Student Engagement in Blended Learning Spaces
Öncü, S., Guillén‐Gámez, F. D., & Çiğdem, H. (2026) · Springer eBooks · doi:10.1007/978-981-95-4871-2_6 · Cited by 2
A Methodological Framework for Venture Capital Investment Decision Making Using Hybrid MCDM and Machine Learning
Kellekci, M., Ufuk ISTANBUL TECHNICAL UNIVERSITY CEBECI, & Doğan, O. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-7416-1.ch005 · Cited by 2
Ethical considerations and challenges in pediatric surgical artificial intelligence
González, R. (2026) · World Journal of Pediatric Surgery · doi:10.1136/wjps-2025-001102 · Cited by 2
Beyond metrics to methods: a scoping review of transformers and large language models for detection of social drivers of health in clinical notes
Farrag, A. N., Soliman, A., Hatef, E., Goodin, A. J., & Rouhizadeh, M. (2025) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocaf201 · Cited by 3
The RAILE Framework: Guiding Ethical and Effective Use of AI in Higher Education
Rahman, M. M. (2025) · International Conference on Education Technology and Computer · doi:10.1109/icetc66579.2025.11387191 · Cited by 2
“Many educators consider these tools opportunities for students to enhance their diverse learning experiences, particularly for neurodivergent and non-native speakers [10], [19], [13].”
Politically Charged: From Ballot Boxes to Boardrooms and Break Rooms—Leadership at the Intersection of Democracy and Business
Akande, A., Kellner, D. M., Bobek, V., Williams, K. A., Odewale, F., Coker, D. C., Horvat, T., Akande, E. T., Thomas, V. P., Altaf, M., et al. (2026) · Elsevier eBooks · doi:10.1016/b978-0-443-40469-6.00001-0 · Cited by 5
Comparative Analysis of Generative Artificial Intelligence in Evaluating Candidate Attributes for Public Employment: East Versus West
Hsieh, C., Wang, D., He, Y., & Hsieh, D. W. (2026) · Review of Public Personnel Administration · doi:10.1177/0734371x261417717 · Cited by 2
AI ‐Driven Epidemiology: The Next Frontier in Precision Public Health
Prema, S. S., Ganapathy, D., & Shanmugamprema, D. (2026) · Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery · doi:10.1002/widm.70077 · Cited by 3
“If these biases are not identified and corrected, AI-generated predictions and recommendations may inadvertently perpetuate health inequities, disproportionately affecting marginalized populations (Hanna et al. 2024; Ferrara 2024).”
KI-Dokumentation in der Psychotherapie
Goer, V., & Schwarz, J. (2026) · Die Psychotherapie · doi:10.1007/s00278-026-00842-4 · Cited by 4
Gender bias in text-to-image generative artificial intelligence: Neglect and stereotypical presentations across three popular platforms
Weinmann, H., Messingschlager, T. V., & Appel, M. (2026) · New Media & Society · doi:10.1177/14614448261435197 · Cited by 3
Explainability, Bias and Generalizability of AI Models in Dentistry: A Systematic Review of Model Interpretability and Equity
Mehta, V., Mathur, A. S., Bhadania, M., Galletti, C., & Fraile, J. F. (2026) · Clinical and Experimental Dental Research · doi:10.1002/cre2.70375 · Cited by 5
“This observation was made by Ferrara (2023), highlighting that the robustness of AI models is affected by their algorithmic structure and the quality of the dataset rather than their qualitative characteristics.”
Is AI-Produced Humanities Scholarship a Case of Research Misconduct?
Metcalf, T. R. (2026) · Journal of Academic Ethics · doi:10.1007/s10805-026-09745-0 · Cited by 3
Exploring Human-Machine Dynamics: How Do Users Understand and Adapt to the Moral Risks in Healthcare Generative AI Chatbots?
Yang, Y., Qi, J., & Ji, X. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2683905 · Cited by 2
Research challenges and emerging futuristic evolution for 3D medical image processing
Upadhyaya, V., & Gupta, N. K. (2024) · Advances in computers · doi:10.1016/bs.adcom.2024.03.004 · Cited by 2
Building Trust and Credibility
Demir, M. (2024) · Advances in hospitality, tourism and the services industry (AHTSI) book series · doi:10.4018/979-8-3693-1239-1.ch010 · Cited by 2
Artificial Intelligence in Skill Development
Bhowmick, M., Goswami, S., Chakraborty, R., Bhowmick, P., Nandy, S. K., Afrose, N., & Kewatkar, S. M. (2024) · Advances in human resources management and organizational development book series · doi:10.4018/979-8-3693-6412-3.ch006 · Cited by 2
Assessing Gender and Racial Bias in Large Language Model‐Powered Virtual Reference
Liu, J., & Wang, H. (2024) · Proceedings of the Association for Information Science and Technology · doi:10.1002/pra2.1061 · Cited by 4
“Prior studies underscore the growing concerns regarding the reliability and bias of generative AI outputs (Ferrara, 2024; Kotek et al., 2023; Tang et al., 2023).”
Artificial intelligence for laryngoscopy in vocal fold diseases: a review of dataset, technology, and ethics
Dao, T. T. P., Nguyen, T.-C., Huynh, V.-T., Bui, X., Le, T.-N., & Tran, M. (2024) · Machine Learning · doi:10.1007/s10994-024-06602-2 · Cited by 3
“Learning from biased data, the model perpetuates these same biases in its predictions, resulting in skewed outcomes (Ferrara, 2023).”
AI-Driven Wildfire Prediction and Response in Portugal
Gomes, P. P., & Cardoso, T. J. (2025) · Advances in environmental engineering and green technologies book series · doi:10.4018/979-8-3693-7483-2.ch005 · Cited by 2
Navigating New Frontiers
Roberts, L. N. (2025) · Advances in educational technologies and instructional design book series · doi:10.4018/979-8-3693-6527-4.ch011 · Cited by 2
Artificial Intelligence in the Stock Market: The Trends and Challenges Regarding AI-Driven Investments
Kotecha, N. (2025) · Open Journal of Business and Management · doi:10.4236/ojbm.2025.132037 · Cited by 2
AI in Corporate Governance: Perspectives from the MENA
Azoury, N., & Hajj, C. E. (2025) · Springer eBooks · doi:10.1007/978-3-031-75589-7_18 · Cited by 2
Gender Disparities in Melanoma: Advances in Diagnosis, Treatment, and the Role of Artificial Intelligence
Kakish, D. R. K., AlSamhori, J. F., Qaqish, L. N., Aburumman, L., Sarsur, R., Salkhadi, A. A., Nassif, Z. B., Akmal, M. A., & Nashwan, A. J. (2025) · Dermatological Reviews · doi:10.1002/der2.70022 · Cited by 3
Strategic Decision-Making in the AI Era: An Integrated Approach Classical, Adaptive, Resource-Based, and Processual Views
Büber, H., & Seven, E. (2025) · International Journal of Management and Administration · doi:10.29064/ijma.1637935 · Cited by 9
WIP: Impact of Generative AI on Learning: Foci and Parameters for an Assessment Tool
Khan, M. M., Ford, C., Dong, Y., & Afsarimanesh, N. (2024) · IEEE Frontiers in Education Conference (FIE) · doi:10.1109/fie61694.2024.10893560 · Cited by 3
Deep Learning Approach for Evaluating Fairness in LLMs
Kumari, L., Serasiya, S., & Bharti, S. K. (2025) · International Conference on Sustainable Energy Technologies and Computational Intelligence (SETCOM) · doi:10.1109/setcom64758.2025.10932507 · Cited by 2
“…have been done on this topic, they have covered more broad topics like deep learning, artificial intelligence or machine learning, they have evaluated bias for specific scenarios and application domains or they have taken LLMs as a whole without taking into account the sizes [9,10,11,12,13,14,15].”
Digital Narratives and AI in Modern Media: A Systematic Literature Review
Harahap, K., Muhtar, Sakarina, S., Shofwani, S. A., & Hariyadi, A. (2025) · Studies in systems, decision and control · doi:10.1007/978-3-031-84636-6_16 · Cited by 1
Innovative Question Types and Formats
Trajkovski, G., & Hayes, H. (2025) · Digital education and learning · doi:10.1007/978-3-031-88252-4_3 · Cited by 2
AI Driven Threat Detection for Autonomous Robots
Taj, S., Khan, Z. U., Amin, S. U., et al. (2025) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-0583-7.ch004 · Cited by 2
La inteligencia artificial generativa en la publicación científica y el papel de los editores en revistas de biblioteconomía, ciencias de la información y educación en Hispanoamérica
Ciro, J. B., & Franco-Rico, J. (2025) · INTER DISCIPLINA · doi:10.22201/ceiich.24485705e.2025.36.90833 · Cited by 2
Role and Use of Race in Artificial Intelligence and Machine Learning Models Related to Health
Were, M. C., Li, A., Malin, B., Yin, Z., Coco, J., Collins, B. X., Clayton, E. W., Novak, L. L., Hendricks‐Sturrup, R., Oluyomi, A., et al. (2025) · Journal of Medical Internet Research · doi:10.2196/73996 · Cited by 4
“…of benefiting from the model), counterfactual fairness (ie, the patient-level model outcomes are unaffected by variations in protected attributes such as race and other demographic information), and group fairness (ie, model outcomes are similar across groups of sensitive attributes) [72].”
Workflow based on GANs and CNNs towards a digital twin for the 3D morphological characterization of latex aggregates
Théodon, L., Coufort‐Saudejaud, C., & Debayle, J. (2025) · Powder Technology · doi:10.1016/j.powtec.2025.121286 · Cited by 4
Harnessing Generative AI for Personalized Learning: A Review in the Context of National Education Policy, 2020
Kadam, M. A., Sindhu, D. P. M., & Joshi, D. S. (2025) · International Journal of Environmental Sciences · doi:10.64252/gnbkf548 · Cited by 2
Automated Bias Assessment in AI-Generated Educational Content Using CEAT Framework
Peng, J., Shen, W., Rao, J., & Lin, J. (2025) · Communications in computer and information science · doi:10.1007/978-3-031-99264-3_44 · Cited by 4
Educational Chatbots: A Systematic Mapping Study
Urquiza-Yllescas, J. F., Mendoza, S., Sánchez-Adame, L. M., Rodríguez, J., & Decouchant, D. (2025) · Journal of Intelligent & Fuzzy Systems · doi:10.1177/18758967251356870 · Cited by 5
“…generative AI models are exacerbated by the data that feeds them, the feedback they learn from, and the input from their developers, leading to skewed representations of various demographics such as ethnicity, sexual orientation, and political views (Ferrara, 2024; Gross, 2023; Huang & Rust, 2021).”
Artificial intelligence in engineering education research: Using machine learning models to predict undergraduate engineering students' persistence to graduation
Osunbunmi, I. S., Feyijimi, T. R., Cutler, S. L., Brijmohan, Y., Arinze, L., Dansu, V., Bamidele, B., Wu, J., & Rabb, R. (2025) · Journal of Engineering Education · doi:10.1002/jee.70034 · Cited by 2
Bio-Centered Artificial Intelligence: Towards a Progressive AI-biosphere Relation Through the Concepts of Poiesis and Mimesis
Stellinga, L., Korenhof, P., & Blok, V. (2025) · Topoi · doi:10.1007/s11245-025-10270-9 · Cited by 2
Securing Machine Learning: Understanding Adversarial Attacks and Bias Mitigation
Lakhani, A., & Rohit, N. (2024) · International Journal of Innovative Science and Research Technology (IJISRT) · doi:10.38124/ijisrt/ijisrt24jun1671 · Cited by 3
Future Approach Generative AI, Stylized Architecture, and its Potential in Finance
Bhatnagar, A. B., Nanad, A., Kumar, A., & Kumar, R. (2025) · Wiley eBooks · doi:10.1002/9781394271078.ch3 · Cited by 3
Artificial Intelligence Technologies and Practical Normativity/Normality: Investigating Practices beyond the Public Space
Bode, I., & Huelss, H. (2024) · Open Research Europe · doi:10.12688/openreseurope.16536.2 · Cited by 2
“Many AI researchers argue that it is possible to circumvent such bias in the training data, inspiring an increasing research stream on bias mitigation strategies and their potential trade-offs (Ferrara, 2023).”

Authors: Ferrara, E.
Affiliations: University of Southern California (US)
Corresponding author: Ferrara, Emilio
ORCID: Ferrara, Emilio
Venue: Sci
Publisher: MDPI AG
Published: 2023-12-26
Volume: 6
Issue: 1
Pages: 3
ISSN: 2413-4155 (ISSN-L)
DOI: 10.3390/sci6010003
OpenAlex: W4390229867
Semantic Scholar: 53b04ccd2a001467d7ce168e9ce20b16a9466a69
Cited by: 875
References: 59
Citations by year: 2026: 271 · 2025: 433 · 2024: 162 · 2023: 6
FWCI: 218.36
Type: journal article
Language: en
Open access: gold
doi.org
Licence: cc-by
creativecommons.org/licenses/by/4.0/ (vor)
Locations: Sci (publishedVersion, cc-by)
SSRN Electronic Journal (acceptedVersion)
arXiv (Cornell University) (submittedVersion, cc-by)
DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
arXiv (Cornell University) (cc-by)
Topics: Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
Social Sciences › Business, Management and Accounting › Organizational Behavior and Human Resource Management › AI and HR Technologies
Keywords: algorithmic fairness, algorithmic bias, generative AI bias, generative AI, synthetic media, data bias, bias mitigation, bias mitigation strategies, data preprocessing, model selection, post-processing, societal impact of AI, stereotypes, artificial intelligence in healthcare, AI ethics

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies

Abstract

Overview

AI snapshot

TL;DR

Key findings

Methods

Limitations

Integrity signals

Verification

Claims the paper makes

Claims other papers make about this one

Self-check

Citations

Cites

Cited by

Related

Metadata

Record

About these checks