Algorithmic Political Bias in Artificial Intelligence Systems
Open accessThe DOI matches its Crossref record; 81 of the 81 references it cites were found in a registry or an index.
Abstract
Some artificial intelligence (AI) systems can display algorithmic bias, i.e. they may produce outputs that unfairly discriminate against people based on their social identity. Much research on this topic focuses on algorithmic bias that disadvantages people based on their gender or racial identity. The related ethical problems are significant and well known. Algorithmic bias against other aspects of people's social identity, for instance, their political orientation, remains largely unexplored. This paper argues that algorithmic bias against people's political orientation can arise in some of the same ways in which algorithmic gender and racial biases emerge. However, it differs importantly from them because there are (in a democratic society) strong social norms against gender and racial biases. This does not hold to the same extent for political biases. Political biases can thus more powerfully influence people, which increases the chances that these biases become embedded in algorithms and makes algorithmic political biases harder to detect and eradicate than gender and racial biases even though they all can produce similar harm. Since some algorithms can now also easily identify people's political orientations against their will, these problems are exacerbated. Algorithmic political bias thus raises substantial and distinctive risks that the AI community should be aware of and examine.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“But, this approach would fail to ensure social neutrality, since a convenience sample of annotators would have unrepresentative biased views on what constitutes immoral actions or generally undesirable outputs (156).”
From The impact of generative artificial intelligence on socioeconomic inequalities and policy making (2024) · cites this paper as [156]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“That most models produced overgeneralizations reliably across time suggests that they may have incorporated an ‘algorithmic bias’ from the training data similar to the way models trained on linguistic data can acquire gender, racial, or political biases, by picking up human tendencies implicit in training text corpora [51–53].”
From Generalization bias in large language model summarization of scientific research (2025) · cites this paper as [53]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
A Survey on Bias and Fairness in Machine Learning
Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., & Galstyan, A. (2019) · arXiv · doi:10.48550/arxiv.1908.09635 · Cited by 5,812
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
Artificial intelligence in healthcare: past, present and future
Jiang, F., Jiang, Y., Zhi, H., Dong, Y., Li, H., Ma, S., Wang, Y., Dong, Q., Shen, H., & Wang, Y. (2017) · Stroke and Vascular Neurology · doi:10.1136/svn-2017-000101 · Cited by 4,837 · cited 2×
“, 2016), or what medical diagnosis a patient should receive (Jiang et al., 2017).”
BackgroundThe Origins and Consequences of Affective Polarization in the United States
Iyengar, S., Lelkes, Y., Levendusky, M., Malhotra, N., & Westwood, S. J. (2019) · Annual Review of Political Science · doi:10.1146/annurev-polisci-051117-073034 · Cited by 3,603 · cited 5×
“Iyengar et al. (2019) suggest that political biases are more pronounced because ‘[u] nlike race, gender, and other social divides where group-related attitudes and behaviors are subject to social norms […], there are [in the USA] no corresponding pressures to temper disapproval of political…”
BackgroundInfluential(2018) · Cited by 3,510
How the machine “thinks”: Understanding opacity in machine learning algorithms
Burrell, J. (2016) · Big Data & Society · doi:10.1177/2053951715622512 · Cited by 3,196 · cited 1×
“The here relevant AI algorithms are mining data according to models that they have formed through machine learning (ML) (Burrell, 2016).”
MethodBarocas, S., & Selbst, A. D. (2014) · doi:10.2139/ssrn.2477899 · Cited by 3,048 · cited 2×
“…features as predictors of low hiring success and disfavour certain gender, racial, or political groups with credentials otherwise equal to other applicants even if no employer ever explicitly indicated the problematic predictor-outcome relationships in previous decisions (Barocas & Selbst, 2016).”
BackgroundAlgorithms of Oppression: How Search Engines Reinforce Racism
Noble, S. U. (2020) · doi:10.2307/j.ctt1pwt9w5 · Cited by 2,709
Fear and loathing across party lines: New evidence on group polarization
Iyengar, S., & Westwood, S. J. (2015) · American Journal of Political Science · doi:10.1111/ajps.12152 · Cited by 2,452
Political Ideology: Its Structure, Functions, and Elective Affinities
Jost, J. T., Federico, C. M., & Napier, J. L. (2009) · Annual Review of Psychology · doi:10.1146/annurev.psych.60.110707.163600 · Cited by 2,125 · cited 4×
“their identity as liberal, conservative, moderate, Marxist, anarchist, and so on (Jost et al., 2009).”
BackgroundInfluentialBellamy, R. K. E., Dey, K., Hind, M., Hoffman, S. C., Houde, S., Kannan, K., Lohia, P., Martino, J., Mehta, S., Mojsilovic, A., et al. (2018) · arXiv · doi:10.48550/arxiv.1810.01943 · Cited by 1,581 · cited 1×
“…contains many concrete recommendations on how to reduce algorithmic gender and racial biases, including interventions that use debiasing algorithms (for data pre-processing before training, in-processing during training, or post-processing after training; Amini et al., 2019; Bellamy et al., 2019).”
BackgroundFairness definitions explained
Verma, S., & Rubin, J. (2018) · Proceedings of the International Workshop on Software Fairness · doi:10.1145/3194770.3194776 · Cited by 1,124 · cited 1×
“There is currently ‘no clear agreement [in the AI community] on which definition [of ‘fairness’] to apply in each situation’ (Verma & Rubin, 2018, p. 1).”
BackgroundFoundations of Machine Learning
Singh, A. (2019) · SSRN Electronic Journal · doi:10.2139/ssrn.3399990 · Cited by 1,123 · cited 1×
“Their accuracy is often related to the size of the datasets that they are trained on: a smaller training dataset commonly produces more inaccuracy, making groups underrepresented in datasets more vulnerable to classification errors (Mohri et al., 2018).”
BackgroundComputer-based personality judgments are more accurate than those made by humans
Youyou, W., Kosinski, M., & Stillwell, D. (2015) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1418680112 · Cited by 1,086 · cited 1×
“For instance, some of these algorithms can infer people’s political orientations simply from a set of their Facebook ‘likes’ where these ‘likes’ are not themselves indicative of any particular political view (Youyou et al., 2015).”
BackgroundPolitical Sectarianism in America
Finkel, E. J., Bail, C. A., Cikara, M., Ditto, P. H., Iyengar, S., Klar, S., Mason, L., McGrath, M. C., Nyhan, B., Rand, D. G., et al. (2020) · Science · doi:10.1126/science.abe1715 · Cited by 931 · cited 5×
“In the USA, ‘political tribalism’ (i.e. people’s viewing themselves as belonging to either the group of liberals or the group of conservatives and displaying overt aversion against political out-group members) is at an all-time high and pervasive (Finkel et al., 2020).”
BackgroundInfluentialAI Fairness 360: An extensible toolkit for detecting and mitigating algorithmic bias
Bellamy, R. K. E., Dey, K., Hind, M., Hoffman, S. C., Houde, S., Kannan, K., Lohia, P., Martino, J., Mehta, S., Mojsilović, A., et al. (2019) · IBM Journal of Research and Development · doi:10.1147/jrd.2019.2942287 · Cited by 884
Out of Sight But Not Out of Mind: Managing Invisible Social Identities in the Workplace
Clair, J. A., Beatty, J. E., & Maclean, T. L. (2005) · Academy of Management Review · doi:10.5465/amr.2005.15281431 · Cited by 866 · cited 1×
“In fact, it seems that people can, if they want to, relatively easily conceal their political identity in everyday life and social environments, at the workplace, and so on (resulting in ‘invisible’ diversity) (Clair et al., 2005).”
BackgroundConsumer credit-risk models via machine-learning algorithms
Khandani, A. E., Kim, A. J., & Lo, A. W. (2010) · Journal of Banking & Finance · doi:10.1016/j.jbankfin.2010.06.001 · Cited by 864
Left and right: the significance of a political distinction
(1997) · Choice Reviews Online · doi:10.5860/choice.34-5919 · Cited by 717
The Oxford Handbook of Political Communication
(2017) · Oxford University Press eBooks · doi:10.1093/oxfordhb/9780199793471.001.0001 · Cited by 602
Reiljan, A. (2020) · European Journal of Political Research · doi:10.1111/1475-6765.12351 · Cited by 592 · cited 1×
“In fact, some studies found that ‘affective [political] polarization is [also] acutely present in European party systems, as partisans are often extremely hostile towards competing parties’ (Reiljan, 2020, p. 1).”
BackgroundHandbook of prejudice, stereotyping, and discrimination
(2010) · Choice Reviews Online · doi:10.5860/choice.47-5949 · Cited by 539
Ditto, P. H., Liu, B. S., Clark, C. J., Wojcik, S. P., Chen, E. E., Grady, R. H., Celniker, J. B., & Zinger, J. F. (2019) · Perspectives on Psychological Science · doi:10.1177/1745691617746796 · Cited by 535 · cited 1×
“Additionally, a recent meta-analysis of studies on political bias found that, for instance, liberal and conservative study participants equally evaluated otherwise identical information more favourably when it supported their prior political identity vs. when it threatened it (Ditto et al., 2019).”
BackgroundKöchling, A., & Wehner, M. C. (2020) · Business Research · doi:10.1007/s40685-020-00134-w · Cited by 511 · cited 1×
“They may then form models that take these factors as predictors and subsequently treat different groups unequally based on gender, race, or political orientation rather than relevant qualification differences (Köchling & Wehner, 2020).”
BackgroundAlgorithmic Bias in Autonomous Systems
Danks, D., & London, A. J. (2017) · Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2017/654 · Cited by 472 · cited 2×
“…neutrally transform or extract information from data but to operate on it in ways that deviate from a normative (moral, statistical, social, etc.) standard such that one kind of individual or group is unfairly privileged over another based on aspects of their social identity (Danks & London, 2017).”
BackgroundAlgorithmic inference, political interest, and exposure to news and politics on Facebook
Thorson, K., Cotter, K., Medeiros, M., & Pak, C. (2021) · Information, Communication & Society · doi:10.1080/1369118x.2019.1642934 · Cited by 361 · cited 1×
“…for relevant information for a given individual (Robertson et al., 2018) and be used by personalization algorithms to selectively present some contents (or people and groups) to that individual and ignore others because of their political orientation (Le et al., 2019; Thorson et al., 2021).”
BackgroundPolitical Diversity in Social and Personality Psychology
Inbar, Y., & Lammers, J. (2012) · Perspectives on Psychological Science · doi:10.1177/1745691612448792 · Cited by 359 · cited 2×
“Yet, surveys with US (Yancey, 2011; Shields & Dunn 2016) and international samples (Inbar & Lammers, 2012; Peters et al., 2020) found that many academics in the sciences and humanities openly expressed willingness to discriminate against colleagues with a political orientation opposite to their own.”
BackgroundCross-Country Trends in Affective Polarization
Boxell, L., Gentzkow, M., & Shapiro, J. (2020) · doi:10.3386/w26669 · Cited by 359
The tie that divides: Cross‐national evidence of the primacy of partyism
Westwood, S. J., Iyengar, S., Walgrave, S., Leonisio, R., Miller, L., & Strijbis, O. (2018) · European Journal of Political Research · doi:10.1111/1475-6765.12228 · Cited by 347 · cited 2×
“First, there are international differences with respect to norms against political biases, discrimination, and polarizations (Boxell et al., 2020; Westwood et al., 2018).”
BackgroundThe Limits of Partisan Prejudice
Lelkes, Y., & Westwood, S. J. (2017) · The Journal of Politics · doi:10.1086/688223 · Cited by 311
Koch, A., Imhoff, R., Dotsch, R., Unkelbach, C., & Alves, H. (2016) · Journal of Personality and Social Psychology · doi:10.1037/pspa0000046 · Cited by 309 · cited 3×
“…whether individuals we meet are ‘engines of change or preservers of the status quo’, and so whether they are politically liberal vs. conservative, is one of the most frequently invoked and ‘fundamental dimensions on which [we] spontaneously distinguish social groups’ (Koch et al., 2016, p. 702).”
BackgroundInfluentialFletcher, R. R., Nakeshimana, A., & Olubeko, O. (2021) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2020.561802 · Cited by 305 · cited 2×
“The meaning of ‘unfair’ and ‘fair’ in the context of algorithms is commonly ‘judged against a set of legal or ethical principles, which tends to vary depending on the local government and culture’ (Fletcher et al., 2021, p. 7).”
BackgroundDoes Politics Influence Hiring? Evidence from a Randomized Experiment
Gift, K., & Gift, T. (2015) · Political Behavior · doi:10.1007/s11109-014-9286-0 · Cited by 293 · cited 1×
“…either a Democrat or Republican identity cue), most participants selected the student with their own political orientation even when the student with the opposite orientation had higher grades (for more evidence of political bias in hiring contexts, see Gift and Gift (2015) and Roth et al. (2020)).”
BackgroundAlgorithmic bias: Senses, sources, solutions
Fazelpour, S., & Danks, D. (2021) · Philosophy Compass · doi:10.1111/phc3.12760 · Cited by 272 · cited 1×
“There has been a flurry of research on algorithmic bias (Fazelpour & Danks, 2021).”
BackgroundKrahé, B., Möller, I., Huesmann, L. R., Kirwil, L., Felber, J., & Berger, A. (2011) · Journal of Personality and Social Psychology · doi:10.1037/a0021711 · Cited by 236 · cited 1×
“This is because increased exposure to these social factors (just as increased exposure to media violence) may desensitize people and increase their tolerance threshold for them (Krahé et al., 2011) .”
BackgroundAlgorithmic bias: on the implicit biases of social technology
Johnson, G. M. (2021) · Synthese · doi:10.1007/s11229-020-02696-y · Cited by 232 · cited 1×
“By treating them as proxies for the sensitive ones, they can produce the same outputs as before (see also the ‘proxy problem’ in Johnson, 2021).”
BackgroundOverdoing Democracy: Why We Must Put Politics in its Place
Talisse, R. B. (2019) · Cited by 213
When Passionate Advocates Meet Research on Diversity, Does the Honest Broker Stand a Chance?
Eagly, A. H. (2016) · Journal of Social Issues · doi:10.1111/josi.12163 · Cited by 209 · cited 1×
“Additionally, if political polarization is widespread and persistent within society itself, political diversity measures in AI teams may result in social conflicts undermining the cooperation of team members (Eagly, 2016).”
BackgroundUncovering and Mitigating Algorithmic Bias through Learned Latent Structure
Amini, A., Soleimany, A. P., Schwarting, W., Bhatia, S. N., & Rus, D. (2019) · Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3306618.3314243 · Cited by 206 · cited 2×
“The literature contains many concrete recommendations on how to reduce algorithmic gender and racial biases, including interventions that use debiasing algorithms (for data pre-processing before training, in-processing during training, or post-processing after training; Amini et al., 2019; Bellamy et al., 2019).”
BackgroundTracking the Digital Footprints of Personality
Lambiotte, R., & Kosinski, M. (2014) · Proceedings of the IEEE · doi:10.1109/jproc.2014.2359054 · Cited by 200 · cited 1×
“…er/ artic le/ 21- non-discr imina tion 9 https:// www. gov. uk/ dismi ss-staff/ unfair-dismi ssals 2019), from their ‘digital footprints’ (e.g. their clicks, news browsing, etc.) even when website users do not explicitly express any political statements (Lambiotte & Kosinski, 2014; Vincent, 2016).”
BackgroundFacial recognition technology can expose political orientation from naturalistic facial images
Kosinski, M. (2021) · Scientific Reports · doi:10.1038/s41598-020-79310-1 · Cited by 199 · cited 3×
“A recent study published in Science (Kosinski, 2021) found that some existing facial recognition algorithms (used by e.”
BackgroundDruckman, J. N., Klar, S., Krupnikov, Y., Levendusky, M., & Ryan, J. B. (2021) · Journal of Experimental Political Science · doi:10.1017/xps.2020.28 · Cited by 193 · cited 1×
“…it is both common (think of Brex-iteers vs. Remainers, populists (anti-immigration) vs. cosmopolitans (pro-immigration advocates), etc., Druckman et al., 2020) and relatively stable across various types of democratic countries, including ones with many different political parties and…”
BackgroundDoes “ AI ” stand for augmenting inequality in the era of covid-19 healthcare?
Leslie, D., Mazumder, A., Peppin, A., Wolters, M. K., & Hagerty, A. (2021) · BMJ · doi:10.1136/bmj.n304 · Cited by 177 · cited 1×
“…records that reflect disparities in healthcare access and quality’, resulting in an AI system that is likely to more frequently mislabel individuals not belonging to the dominant group and so will ‘reflect, repeat, and compound pre-existing structural discrimination’ (Leslie et al., 2021, p. 2).”
BackgroundThe age of secrecy and unfairness in recidivism prediction
Rudin, C., Wang, C., & Coker, B. (2020) · Harvard Data Science Review · doi:10.1162/99608f92.6ed64b30 · Cited by 177
On the Left and Right Ideological Divide: Historical Accounts and Contemporary Perspectives
Caprara, G. V., & Vecchione, M. (2018) · Political Psychology · doi:10.1111/pops.12476 · Cited by 170 · cited 1×
“…internationalism, and state intervention to regulate the economy, whereas the ‘right is invariably identi-fied’ with traditional values, authority, order, nationalism, ‘market liberalization, and lesser state intervention’ in the economy (Rosas & Ferreira, 2013, p. 9; Caprara & Vecchione, 2018).”
BackgroundBiased Programmers? Or Biased Data? A Field Experiment in Operationalizing AI Ethics
Cowgill, B., Dell'Acqua, F., Deng, S., Hsu, D., Verma, N., & Chaintreau, A. (2020) · Proceedings of the 21st ACM Conference on Economics and Computation · doi:10.1145/3391403.3399545 · Cited by 166 · cited 1×
“The more heterogeneous the teams developing AI algorithms and researching their implications, the higher the likelihood that biases in data selection, labelling, and programming are detected and counteracted (Cowgill et al., 2020).”
BackgroundNOISE: How to overcome the high, hidden cost of inconsistent decision making
Kahneman, D. J., Rosenfield, A. M., Gandhi, L., & Blaser, T. (2016) · Harvard business review · Cited by 156
My enemy's enemy is my friend.
Aronson, E., & Cope, V. Z. (1968) · Journal of Personality and Social Psychology · doi:10.1037/h0021234 · Cited by 154 · cited 1×
“Underestimations may be especially common amongst people whose political opponents are targeted by these biases, as the power of the rationale that the ‘enemy of my enemy is my friend’ is well documented (Aronson & Cope, 1968).”
BackgroundGenetic and Environmental Transmission of Political Orientations
Funk, C. L., Smith, K. B., Alford, J. R., Hibbing, M. V., Eaton, N. R., Krueger, R. F., Eaves, L. J., & Hibbing, J. R. (2012) · Political Psychology · doi:10.1111/j.1467-9221.2012.00915.x · Cited by 153 · cited 1×
“However, there is evidence that political orientation is not always fully chosen but partly biologically determined (Funk et al., 2013; Kalmoe & Johnson, 2021; Tilley, 2021).”
BackgroundKozyreva, A., Lorenz-Spreen, P., Hertwig, R., Lewandowsky, S., & Herzog, S. M. (2021) · Humanities and Social Sciences Communications · doi:10.1057/s41599-021-00787-w · Cited by 150 · cited 1×
“First, many websites employ personalization algorithms to provide web-site users with content similar to what they previously viewed so as to keep them engaged (Kozyreva et al., 2021).”
BackgroundHuman subjects review, personal values, and the regulation of social science research.
Ceci, S. J., Peters, D. P., & Plotkin, J. (1985) · American Psychologist · doi:10.1037/0003-066x.40.9.994 · Cited by 146 · cited 1×
“Studies in which political biases affected decisions on scholarship award (Iyengar & Westwood, 2015), job hiring (Roth et al., 2020), or research manuscripts (Abramowitz et al., 1975; Ceci et al., 1985) illustrate this.”
BackgroundAlgorithmic Censorship by Social Platforms: Power and Resistance
Cobbe, J. (2021) · Philosophy & Technology · doi:10.1007/s13347-020-00429-0 · Cited by 146 · cited 1×
“Similarly, social media algorithms may be trained to proactively block or remove some political, dangerous, or untrustworthy information (e.g. Nazi content) (Cobbe, 2020).”
BackgroundJohndrow, J. E., & Lum, K. (2019) · The Annals of Applied Statistics · doi:10.1214/18-aoas1201 · Cited by 141
Auditing the Personalization and Composition of Politically-Related Search Engine Results Pages
Robertson, R. E., Lazer, D., & Wilson, C. (2018) · Proceedings of the 2018 World Wide Web Conference on World Wide Web - WWW '18 · doi:10.1145/3178876.3186143 · Cited by 139 · cited 2×
“Markers of political orientation might become predictors for relevant information for a given individual (Robertson et al., 2018) and be used by personalization algorithms to selectively present some contents (or people and groups) to that individual and ignore others because of their political…”
BackgroundPublish or Politic: Referee Bias in Manuscript Review1
Abramowitz, S. I., Gomes, B., & Abramowitz, C. V. (1975) · Journal of Applied Social Psychology · doi:10.1111/j.1559-1816.1975.tb00675.x · Cited by 137 · cited 1×
“Studies in which political biases affected decisions on scholarship award (Iyengar & Westwood, 2015), job hiring (Roth et al., 2020), or research manuscripts (Abramowitz et al., 1975; Ceci et al., 1985) illustrate this.”
BackgroundForecasting Domestic Violence: A Machine Learning Approach to Help Inform Arraignment Decisions
Berk, R. A., Sorenson, S. B., & Barnes, G. (2016) · Journal of Empirical Legal Studies · doi:10.1111/jels.12098 · Cited by 133 · cited 1×
“…to decide, for instance, whether a job applicant is suitable for a vacancy (Bogen, 2019), whether someone is eligible to receive a loan (Khandani et al., 2010), how long a convict should stay incarcerated (Berk et al., 2016), or what medical diagnosis a patient should receive (Jiang et al., 2017).”
BackgroundHuman and Computer Personality Prediction From Digital Footprints
Hinds, J., & Joinson, A. (2019) · Current Directions in Psychological Science · doi:10.1177/0963721419827849 · Cited by 129
Broockman, D., Ferenstein, G., & Malhotra, N. A. (2018) · American Journal of Political Science · doi:10.1111/ajps.12408 · Cited by 122 · cited 3×
“Similarly, if a hiring algorithm for AI job positions is trained on recruitment data only from Silicon Valley, since people on the radical left and conservatives tend to be underrepresented in this area (Broockman et al., 2019; Tiku, 2018), the algorithm may subsequently downgrade their AI job applications, as their p…”
BackgroundMethodInfluentialCompromising Scholarship: Religious and Political Bias in American Higher Education
Yancey, G. (2011) · ScholarsArchive (Brigham Young University) · Cited by 122
Crawford, J. T., Brandt, M. J., Inbar, Y., Chambers, J. R., & Motyl, M. (2017) · Journal of Personality and Social Psychology · doi:10.1037/pspa0000074 · Cited by 114 · cited 2×
“Some positions on the left and right might overlap (Crawford et al., 2017), and both sit on a spectrum containing many different positions ranging from slightly left- or right-leaning to extremely left- or right-leaning viewpoints (Heywood, 2015).”
BackgroundEkström, M., Patrona, M., & Thornborrow, J. (2020) · Social Semiotics · doi:10.1080/10350330.2020.1762984 · Cited by 108 · cited 1×
“Political views are on a spectrum between extremes, where setting a particular point as a ‘red line’ can become arbitrary, hotly debated, or based on prior value judgements that change over time (Ekström et al., 2020).”
BackgroundGross, N. (2014) · Johns Hopkins University Press eBooks · doi:10.1353/book.31449 · Cited by 104 · cited 2×
“Consider, then, the finding that in US academia, overall, ‘Marxists are rare’ (< 18% of all pro-fessors; Gross & Simmons, 2014, p. 33).”
BackgroundShields, J. A., & Dunn, J. M. (2016) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780199863051.001.0001 · Cited by 102
Roth, P. L., Thatcher, J. B., Bobko, P., Matthews, K. D., Ellingson, J. E., & Goldberg, C. (2020) · Journal of Applied Psychology · doi:10.1037/apl0000422 · Cited by 100 · cited 8×
“…is some evidence that job-related information such as knowledge and skills can attenuate the effect of ethnicity and gender on hiring decisions, Roth et al. (2020) found that ‘political similarity processes continued to significantly influence hire-ability ratings even when information about…”
BackgroundInfluentialThe race to the top among the world’s leaders in artificial intelligence
Savage, N. (2020) · Nature · doi:10.1038/d41586-020-03409-8 · Cited by 87 · cited 7×
“In the USA, ‘political tribalism’ (i.e. people’s viewing themselves as belonging to either the group of liberals or the group of conservatives and displaying overt aversion against political out-group members) is at an all-time high and pervasive (Finkel et al., 2020).”
BackgroundInfluentialLi, D., Raymond, L., & Bergman, P. (2020) · doi:10.3386/w27736 · Cited by 64
Key Concepts in Politics and International Relations
Heywood, A. (2015) · doi:10.1007/978-1-137-49477-1 · Cited by 56 · cited 4×
“Some positions on the left and right might overlap (Crawford et al., 2017), and both sit on a spectrum containing many different positions ranging from slightly left-or right-leaning to extremely left-or right-leaning viewpoints (Heywood, 2015).”
BackgroundInfluentialThe Political Uses and Abuses of Civility and Incivility
Jamieson, K. H., Volinsky, A. C., Weitz, I., & Kenski, K. M. (2015) · Oxford University Press eBooks · doi:10.1093/oxfordhb/9780199793471.013.79 · Cited by 53
Taming The Golem: Challenges of Ethical Algorithmic Decision-Making
Tene, O., & Polonetsky, J. (2017) · North Carolina journal of law & technology · Cited by 52
Ideological diversity, hostility, and discrimination in philosophy
Peters, U., Honeycutt, N., De Block, A., & Jussim, L. (2020) · Philosophical Psychology · doi:10.1080/09515089.2020.1743257 · Cited by 51 · cited 2×
“Yet, surveys with US (Yancey, 2011; Shields & Dunn 2016) and international samples (Inbar & Lammers, 2012; Peters et al., 2020) found that many academics in the sciences and humanities openly expressed willingness to discriminate against colleagues with a political orientation opposite to their own.”
BackgroundSelf-citationBald and Bad? Experimental Evidence for a Dual-Process Account of Baldness Stereotyping
Kranz, D., Nadarevic, L., & Erdfelder, E. (2019) · Experimental Psychology (formerly Zeitschrift für Experimentelle Psychologie) · doi:10.1027/1618-3169/a000457 · Cited by 41 · cited 1×
“While hair colour (Stollznow, 2021), baldness (Kranz et al., 2019), or wearing glasses can be targets of human and algorithmic biases (Seo et al., 2021), such biases are likely much less pronounced and established (Walline et al., 2008) than gender, racial, and political biases.”
BackgroundControlling Prejudice and Stereotyping: Antecedents, Mechanisms, and Contexts
Bodenhausen, G. V., Todd, A. R., & Richeson, J. A. (2009) · Cited by 34
What do kids think about kids in eyeglasses?
Walline, J. J., Sinnott, L. T., Johnson, E. D., Tićak, A., Jones, S. L., & Jones, L. A. (2008) · Ophthalmic and Physiological Optics · doi:10.1111/j.1475-1313.2008.00559.x · Cited by 30 · cited 1×
“While hair colour (Stollznow, 2021), baldness (Kranz et al., 2019), or wearing glasses can be targets of human and algorithmic biases (Seo et al., 2021), such biases are likely much less pronounced and established (Walline et al., 2008) than gender, racial, and political biases.”
BackgroundSantow, E. (2020) · Science · doi:10.1126/science.abb9369 · Cited by 26 · cited 2×
“…study published in Science (Kosin-ski, 2021) found that some existing facial recognition algorithms (used by e.g. the London Metropolitan Police; Santow, 2020) can be trained to ‘expose individuals’ political orientation, as faces of liberals and conservatives consistently differ’: ‘Political…”
They posted what? Recruiter use of social media for selection
Henderson, K. E. (2018) · Organizational Dynamics · doi:10.1016/j.orgdyn.2018.05.005 · Cited by 19
Genes, Ideology, and Sophistication
Kalmoe, N. P., & Johnson, M. (2021) · Journal of Experimental Political Science · doi:10.1017/xps.2021.4 · Cited by 12 · cited 1×
“However, there is evidence that political orientation is not always fully chosen but partly biologically determined (Funk et al., 2013; Kalmoe & Johnson, 2021; Tilley, 2021).”
BackgroundImplicit social cognition: A brief (and gentle) introductionPreprint
Kurdi, B., & Banaji, M. R. (2021) · PsyArXiv (OSF Preprints) · doi:10.31234/osf.io/a4pjy · Cited by 10 · cited 1×
“…the same methodology commonly employed to identify implicit gender and racial biases, i.e. the Implicit Association Test (IAT) (for details, see Kurdi & Banaji, 2021), Iyengar and Westwood (2015) found that Democrats and Republicans showed strong automatic associations between negative words…”
MethodA Characterization of Prediction ErrorsPreprint
Meek, C. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.05955 · Cited by 10 · cited 1×
“false positives, false negatives, generalization errors) and they might arise in different ways (Meek, 2016; Mohri et al., 2018).”
BackgroundLeft and Right the Great Dichotomy Revisited
Rosas, J. C., & Ferreira, A. R. (2013) · Cited by 2
How to disagree, without being disagreeable
(2024) · doi:10.5040/9781399415231.0017 · Cited by 2
Political bias in media doesn’t threaten democracy — other, less visible biases do
Hershey, M. R. (2020) · doi:10.64628/aai.7csxkuqcc · Cited by 1
Research works printed but not linked: 4
- Hagerty, A., & Rubinov, I. (2019). Global AI ethics: A review of the social impacts and ethical implications of artificial intelligence. http://arxiv.org/abs/1907.07892. Accessed 10/04/2020. · arXiv
- Le, H., High, A., Maragh, R., Havens, T., Ekdale, B., & Shafiq, Z. (2019). Measuring political personalization of Google news search. In The Web Conference 2019 - Proceedings of the World Wide Web Conference, WWW 2019 (pp. 2957–2963). (The Web Conference 2019 - Proceedings of the World Wide Web Conference, WWW 2019). Association for Computing Machinery, Inc. 10.1145/3308558.3312504 · doi:10.1145/3308558.3312504
- Noble SU Algorithms of oppression 2019 NYU Press
- Seo, S. Lee, J., & Han, B. (2021). Unsupervised learning of debiased representations with pseudo-attributes. arXiv. https://arxiv.org/abs/2108.02943. Accessed 11/12/2021. · arXiv
Cited by
Habbal, A., Ali, M. K., & Abuzaraida, M. A. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.122442 · Cited by 407
The impact of generative artificial intelligence on socioeconomic inequalities and policy making
Capraro, V., Lentsch, A., Acemoglu, D., Akgun, S., Akhmedova, A., Bilancini, E., Bonnefon, J.-F., Brañas-Garza, P., Butera, L., Douglas, K. M., et al. (2024) · PNAS Nexus · doi:10.1093/pnasnexus/pgae191 · Cited by 371
“Amongst other proposals, the alternative white paper argues that: government use of AI must be transparent, transparency requirements must be mandatory, there must be clear mechanisms for accountability, the public should be consulted about new automated decision-making tools before they are…”
More human than human: measuring ChatGPT political bias
Motoki, F., Pinho Neto, V., & Rodrigues, V. (2024) · Public Choice · doi:10.1007/s11127-023-01097-2 · Cited by 329
“One typical concern would be how AI-powered systems could discriminate against people based on their characteristics, like gender, ethnicity, age, or, more subtly, political beliefs (Peters, 2022).”
InfluentialThe Self-Perception and Political Biases of ChatGPT
Rutinowski, J., Franke, S., Endendyk, J., Dormuth, I., & Pauly, M. (2023) · arXiv · doi:10.48550/arxiv.2304.07333 · Cited by 178
Generalization bias in large language model summarization of scientific research
Peters, U., & Chin-Yee, B. (2025) · Royal Society Open Science · doi:10.1098/rsos.241776 · Cited by 126
Self-citation
Metadata
- Authors
- Peters, U.
- Affiliations
- University of Bonn (DE)
- Corresponding author
- Peters, Uwe
- Venue
- Philosophy & Technology
- Published
- 2022-03-30
- Volume
- 35
- Issue
- 2
- Pages
- 25-25
- ISSN
- 2210-5433 (ISSN-L)
- 2210-5441
- PMC
- PMC8967082
- OpenAlex
- W4221031488
- Cited by
- 111
- References
- 81
- Citations by year
- 2026: 30 · 2025: 30 · 2024: 25 · 2023: 19 · 2022: 7
- FWCI
- 23.49
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0 (tdm)
- creativecommons.org/licenses/by/4.0 (vor)
- Locations
- Philosophy & Technology (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- PhilSci-Archive (University of Pittsburgh) (acceptedVersion)
- PhilPapers (PhilPapers Foundation) (submittedVersion)
- Apollo (University of Cambridge) (submittedVersion, cc-by)
- Apollo
- Funders
- Rheinische Friedrich-Wilhelms-Universität Bonn
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Sociology and Political Science › Feminist Epistemology and Gender Studies
- Social Sciences › Law › Artificial Intelligence in Law
- Keywords
- algorithmic bias, political bias, political orientation, social identity, AI ethics, gender bias, racial bias, social norms, democratic society, political profiling, algorithmic discrimination
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.
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.