Co- unded by he Eu opean Union. The opinions and iews exp essed
a e solely hose o he au ho (s) and do no necessa ily e lec hose
o he Eu opean Union o he Spanish Se ice o he
In e na ionalisa ion o Educa ion (SEPIE). Nei he he Eu opean Union
no he g an ing au ho i y can be held esponsible o hem.
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
„Mind he Gap: Gende Da a and AI Bias“
Gende Da a Gap: E idence B ie –
Suppo ing E idence o OER
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
GEDIS
Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
Ba celona, 05/09/2025
Ci a ion: Bosshamme , S e lana and Daniela Va bano a. (2025). Mind he Gap:
Gende Da a and AI Bias. DOI: 10.5281/zenodo.17164102
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
Execu i e Summa y
The Gende Da a Gap—sys ema ic absence, unde - ep esen a ion, o dis o ion o
da a on women and gi ls—pe pe ua es a male de aul in AI, policy, and p ac ice. This
b ie syn hesizes 13 empi ical cases (2020–2025) ac oss ou bias ypes—
ep esen a i e, algo i hmic, cul u al, in e sec ional—and ou lines key d i e s
(his o ically male-cen ic da a, digi al di ide, me hodological limi s on
in e sec ionali y, unde in es men ). I consolida es ac ionable guidance: de elope s
should di e si y da ase s, in eg a e ai ness mechanisms, and moni o bias;
o ganisa ions should equi e p e/pos -deploymen audi s, es ablish AI e hics
commi ees, and hi e di e sely; egula o s should classi y HR/ inance as high- isk,
manda e gende impac s a emen s, and en o ce algo i hmic anspa ency.
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
Table o con en s
1. De ini ion & F amewo k ................................................................................................................. 5
2. Rele ance & Impac ......................................................................................................................... 5
3. Me hodology ...................................................................................................................................... 6
4. Causes o he Gende Da a Gap.................................................................................................. 7
5. Illus a i e Cases .............................................................................................................................. 8
5.1. Rep esen a i e bias ...................................................................................................................... 8
5.2. Algo i hmic bias: ...................................................................................................................... 10
5.3. Cul u al bias .............................................................................................................................. 12
5.4. In e sec ional bias ................................................................................................................. 13
6. Bes P ac ices and Recommenda ions ................................................................................... 15
6.1. Recommenda ions o AI De elope s .............................................................................. 15
6.2. Recommenda ions o O ganiza ions ...............................................................................16
Re e ences .............................................................................................................................................19
5
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
1. De ini ion & F amewo k
The Gende Da a Gap is he sys ema ic absence, unde - ep esen a ion, o dis o ion
o da a abou women and gi ls ac oss how in o ma ion is collec ed, ep esen ed,
analyzed, and used. Because he e idence base ha in o ms o ganisa ional and
echnological decisions is o en skewed owa d (whi e) men—missing, incomple e, o
lowe -quali y o women— he esul ing models, policies, and p oduc s a e calib a ed
o male bodies, p e e ences, and li e pa hs and ail o cap u e women’s expe iences,
needs, and con ibu ions.
1
,
2
Gende s a is ics a e de ined by da a ha a e: (1) collec ed and p esen ed by sex as
p ima y classi ica ion, (2) e lec gende issues, (3) based on concep s ha
adequa ely cap u e di e si y o women and men's li es, and (4) use collec ion
me hods ha accoun o s e eo ypes and cul u al ac o s ha may induce gende
bias.
3
2. Rele ance & Impac
The Gende Da a Gap skews decisions ac oss socie y and managemen science by
no malizing a male de aul a he han a popula ion- ep esen a i e iew. I has eal
sa e y and heal h cos s, because male-o ien ed da a and aining ma e ials co ela e
wi h wo se ou comes o women. I also slows wo kplace equali y, as policies and
measu emen scales calib a ed on male pa e ns disad an age women and ep oduce
leade ship gaps. This gap unde mines claims o “neu al” managemen heo y, since
1
Spe be e al., "Gende Da a Gap and Managemen Science," 2–8.
2
PARIS21 and UN Women, Gende Da a Ou look 2024.
3
Uni ed Na ions S a is ics Di ision, In eg a ing a Gende Pe spec i e in o S a is ics.
6
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
canonical cons uc s a e o en no med on male beha iou , shi ing a en ion o “ ix-
he-women” a he han s uc u al change. The AI wa e ampli ies hese p oblems:
sys ems ained on male-skewed o un ep esen a i e aces lea n p oxies o gende
and au oma e disc imina ion a scale—making closu e o he gap u gen .
4
A ailabili y does no gua an ee use. Al hough he supply o gende da a has
expanded ac oss many sec o s, up ake emains concen a ed in a ew es ablished
a eas, no ably iolence agains women and unpaid ca e. Many o he domains—
including hose linked o AI—s ill see limi ed applica ion.
To b oaden use, awa eness and access mus imp o e. Dissemina ion should be
pu pose ul, linking p oduce s and use s o s imula e wide up ake and e eal he
needs o new use g oups. Clea , plain-language p esen a ion o gende da a is
essen ial o each he gene al public and s akeholde s wi h lowe da a li e acy.
„An in es men in gende da a is ul ima ely an in es men in he li es o women, gi ls,
boys and men“.
5
3. Me hodology
This e idence b ie employed a sys ema ic app oach o iden i y documen ed cases o
gende bias in AI sys ems. A comp ehensi e li e a u e sea ch was conduc ed ac oss
majo academic da abases (Web o Science, Scopus, PubMed, Google Schola )
co e ing publica ions om 2018 o 2025, using a ge ed keywo ds combining
"a i icial in elligence," "algo i hmic bias," and "gende disc imina ion." F om an ini ial
co pus o o e 30 ele an s udies, 13 ep esen a i e cases we e selec ed based on
empi ical e idence o measu able gende dispa i ies in AI ou comes. Key limi a ions
4
Spe be e al., " Gende Da a Gap and Managemen Science"
5
PARIS21 and UN Women, Gende Da a Ou look 2024
7
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
include ocus on English-language sou ces and comp essed p ojec ime ames ha
limi ed sys ema ic e iew p o ocols.
4. Causes o he Gende Da a Gap
Based on academic esea ch, he Gende Da a Gap s ems om mul iple
in e connec ed ac o s:
• His o ical male-cen ic da a collec ion p ac ices—Sys ema ic p io i iza ion o
men's expe iences, bodies, and li e pa e ns as he de aul no m in esea ch
design and da a ga he ing
6
• Digi al di ide and echnological exclusion—Women's limi ed access o
sma phones, in e ne , and digi al pla o ms educing hei ep esen a ion in
inc easingly impo an digi al da a sou ces
7
• Me hodological challenges in cap u ing in e sec ionali y—Di icul y in designing
esea ch amewo ks ha adequa ely ep esen he di e si y o women's
expe iences ac oss ace, class, disabili y, and o he iden i y ma ke s
8
.
• Ins i u ional unde in es men in gende -speci ic esea ch—Sys ema ic
unde unding o s udies ocused on women's expe iences and gende -
disagg ega ed analysis
9
.
• Biases in esea ch and policy amewo ks—Male-domina ed academic and policy
ins i u ions pe pe ua ing esea ch p io i ies ha e lec masculine
pe spec i es and conce ns
10
.
6
Spe be e al., “Gende Da a Gap and I s Impac on Managemen Science—Re lec ions om a Eu opean Pe spec i e.”
7
Musiz ingoza, “B idging he Gende Da a Gap: Ha nessing Syn he ic Da a o Inclusi e AI,” UNU Macau (blog).
8
Bu inic e al., Mapping Gende Da a Gaps.
9
Ca C iado Pe ez, In isible Women: Exposing Da a Bias in a Wo ld Designed o Men.
10
Spe be e al., “Gende Da a Gap and I s Impac on Managemen Science—Re lec ions om a Eu opean Pe spec i e.”
8
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
5. Illus a i e Cases
We ha e ca ego ized he iden i ied cases in o ou ypes o bias— ep esen a i e, algo i hmic, cul u al, and
in e sec ional—whe e bias e e s o sys ema ic skew in da a, models, o design ha disad an ages ce ain
g oups.
5.1. Rep esen a i e bias
Occu s when da ase s unde - o o e - ep esen speci ic demog aphic g oups, leading
o poo e model pe o mance o hose g oups.
Case 1: Comme cial ace-analysis misclassi ies da ke -skinned women up
o 34.7% s 0.8% o ligh e -skinned men. Face ecogni ion e o s hi
da ke -skinned women ha des , inc easing alse ma ches in policing and
bo de con ol.
An in e sec ional audi o h ee comme cial gende classi ica ion sys ems
(Mic oso , IBM, Face++) e ealed se e e algo i hmic bias. The s udy ound
ha da ke -skinned women we e misclassi ied a a es up o 34.7% compa ed
o jus 0.8% o ligh e -skinned men. Using he newly c ea ed Pilo
Pa liamen s Benchma k da ase balanced by gende and skin ype,
esea che s exposed ha exis ing da ase s we e o e whelmingly composed
o ligh e -skinned subjec s (79.6-86.2%). Despi e comp ising only 21.3% o
he da ase , da ke -skinned emales accoun ed o 61-72% o all classi ica ion
e o s, wi h maximum subg oup dispa i ies eaching 34.4% be ween bes and
wo s classi ied g oups
11
.
11
Buolamwini and Geb u, “Gende Shades: In e sec ional Accu acy Dispa i ies in Comme cial Gende Classi ica ion,” in P oceedings o he 1s
Con e ence on Fai ness, Accoun abili y, and T anspa ency, 77–91.
9
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
Case 2: DALL·E 2 unde ep esen s women (38% s 62% men) and shows
women smiling ~2.2× mo e. In emale-domina ed jobs women a e mo e
o en pic u ed wi h downwa d head pi ch (subo dina ion cue).
An audi o DALL·E 2 (15300 images ac oss 153 US occupa ions) inds bo h
ep esen a ional and p esen a ional gende bias: women appea in only 38.
4% o occupa ional images ( s 46. 4% in Google Images), wi h
unde ep esen a ion in male-domina ed ields and o e ep esen a ion in
emale-domina ed oles; DALL·E 2 ma ches census gende gaps in 88
occupa ions (Google: 18), indica ing ampli ica ion. P esen a ionally, women
a e 2. 19× mo e likely o be shown smiling, and in emale-domina ed jobs mo e
o en depic ed wi h downwa d head pi ch (a subo dina ion cue). These e ec s
exceed hose in Google Images, sugges ing DALL·E 2 no only ep oduces bu
ampli ies occupa ional gende s e eo ypes— h ough bo h who is shown and
how hey a e po ayed.
12
Case 3: AI STEM images o en show 75–100% men, ein o cing “STEM =
male.”
A UNDP Se bia policy analysis o AI image gene a o s inds sys ema ic
unde ep esen a ion o women in STEM: AI-gene a ed STEM images show
men in 75–100% o isuals, despi e women comp ising ~28–40% o STEM
g adua es globally. These sys ems ep oduce—and o en ampli y—
inequali ies: p omp s o “enginee /ma hema ician/scien is ” p edominan ly
yield male igu es, ein o cing “STEM = male.” Such isuals isk sel - ul illing
e ec s by shaping aspi a ions and p o essional iden i y, po en ially lowe ing
12
Sun e al., “Smiling Women Pi ching Down: Audi ing Rep esen a ional and P esen a ional Gende Biases in Image-Gene a i e AI.”
16
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
6.2. Recommenda ions o O ganiza ions
• Manda e p e- and pos -deploymen audi s
29
Conduc sys ema ic echnical and p ocedu al e alua ions be o e and a e AI sys em
deploymen o iden i y and mi iga e eme gen biases. E hical guidelines ad oca e o
egula mul idisciplina y audi s encompassing da a, algo i hms, and use impac s.
• Es ablish AI e hics commi ees
30
Fo m dedica ed in e disciplina y bodies o o e see AI go e nance, in eg a ing
echnical, legal, and social expe ise. Such commi ees ensu e ongoing accoun abili y
and alignmen wi h o ganisa ional alues and e hical s anda ds.
• In es in di e se hi ing p ac ices
31
,
32
,
33
P io i ize ec ui men o p o essionals om a ied demog aphic and disciplina y
backg ounds o enhance eam pe spec i es and educe blind spo s in AI
de elopmen .
6.3. Recommenda ions o Regula o s
• Classi y HR and inancial se ices as high- isk sec o s
34
,
35
,
36
,
37
29
Jobin, Ienca, and Vayena, “The Global Landscape o AI E hics Guidelines.”
30
Jobin, Ienca, and Vayena, “The Global Landscape o AI E hics Guidelines.”
31
Boinodi is, “The Impo ance o AI Di e si y: D i ing T us wo hy AI,” IBM Consul ing.
32
B ad o d, “Why Di e si y in AI Makes Be e AI o All: The Case o Inclusi i y and Inno a ion,” SHRM.
33
Suighi and Rachel, “B eaking he Echo Chambe : Why Di e si y Is C ucial o AI’s Fu u e,” AIM Resea ch Council.
34
Eu opean Union, “AI Ac ,” Shaping Eu ope’s Digi al Fu u e.
35
C isan o e al., “Regula ing AI in he Financial Sec o : Recen De elopmen s and Main Challenges,” Bank o In e na ional Se lemen s (FSI
Insigh s).
36
an de Me we and Veldsman, “AI Risk Managemen o HR: 3 Key Risks To Manage & HR Ac ions To Take,” AIHR.
37
Soleimani e al., “Reducing AI Bias in Rec ui men and Selec ion: An In eg a i e G ounded App oach,” The In e na ional Jou nal o Human
Resou ce Managemen .
17
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
Due o hei di ec impac on employmen , c edi access, and economic oppo uni ies,
AI applica ions in hi ing and lending wa an s ic e o e sigh and manda o y bias
mi iga ion equi emen s.
• Requi e gende impac s a emen s
38
Manda e ha AI de elope s and deploye s publish assessmen s de ailing po en ial
gende - ela ed ha ms, da a gaps, and mi iga ion plans. These s a emen s, g ounded
in human- igh s amewo ks, enable anspa ency and accoun abili y h oughou he
AI li ecycle.
• En o ce algo i hmic anspa ency s anda ds
39
Es ablish legal equi emen s o disclosing AI decision-making logic, key a iables,
and pe o mance me ics. P o iding explainable ou pu s imp o es s akeholde s ’
abili y o de ec and coun e ac bias in c i ical con ex s.
38
Jobin, Ienca, and Vayena, “The Global Landscape o AI E hics Guidelines.”
39
Hou, Tseng, and Yuan, “Is This AI Sexis ? The E ec s o a Biased AI’s An h opomo phic Appea ance and Explainabili y on Use s’ Bias Pe cep ions
and T us ,” In e na ional Jou nal o In o ma ion Managemen .
18
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
Acknowledgemen s — Use o AI ools
We used an AI assis an (Cha GPT, GPT-5 Thinking; Sep 2025) o d a ing suppo
(summa iza ion, copy-edi ing, cla i y). We also used Pe plexi y o li e a u e
disco e y. All sou ces ci ed in his b ie we e loca ed in pee - e iewed o o icial
enues and independen ly e i ied by he au ho s. Gene a i e ou pu s we e no
ea ed as e idence. No pe sonal o con iden ial in o ma ion was en e ed in o hese
ools.
19
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
Re e ences
Spe be , Sonja, Susanne Täube , Co inne Pos , and Co dula Ba zan ny. 2023. "Gende Da a Gap and i s
Impac on Managemen Science — Re lec ions om a Eu opean Pe spec i e." Eu opean
Managemen Jou nal 41, no. 12-8. h ps://doi.o g/10.1016/j.emj.2022.11.006.
PARIS21 and UN Women. Gende Da a Ou look 2024: Unlocking Capaci y, D i ing Change. Pa is:
PARIS21, 2024.
Uni ed Na ions S a is ics Di ision. In eg a ing a Gende Pe spec i e in o S a is ics. New Yo k: Uni ed
Na ions, 2016. Accessed Augus 23, 2025. h ps://uns a s.un.o g/unsd/demog aphic-
social/S anda ds-and-Me hods/ iles/Handbooks/gende /In eg a ing-a-Gende -Pe spec i e-in o-
S a is ics-E.pd .
Ronald, Musiz ingoza. "B idging he Gende Da a Gap: Ha nessing Syn he ic Da a o Inclusi e AI."
UNU Macau (blog). Augus 2, 2024. h ps://unu.edu/macau/blog-pos /b idging-gende -da a-gap-
ha nessing-syn he ic-da a-inclusi e-ai.
Bu inic, May a, Rebecca Fu s -Nichols, and Gaya i Koolwal. Mapping Gende Da a Gaps. Washing on,
DC: Da a2X, Ma ch 2014. h p://da a2x.o g/wp-
con en /uploads/2019/05/Da a2X_MappingGende Da aGaps_FullRepo .pd .
C iado Pe ez, Ca oline. In isible Women: Exposing Da a Bias in a Wo ld Designed o Men. London:
Cha o & Windus, 2019.
Buolamwini, Joy, and Timni Geb u. "Gende Shades: In e sec ional Accu acy Dispa i ies in Comme cial
Gende Classi ica ion." In P oceedings o he 1s Con e ence on Fai ness, Accoun abili y, and
T anspa ency. P oceedings o Machine Lea ning Resea ch 81,
2018. h ps://p oceedings.ml .p ess/ 81/buolamwini18a.h ml.
Chen, Zhisheng. "E hics and Disc imina ion in A i icial In elligence-Enabled Rec ui men
P ac ices." Humani ies and Social Sciences Communica ions 10 (2023):
567. h ps://doi.o g/10.1057/s41599-023-02079-x.
Ta man, Rachael. "Gende and Dialec Bias in YouTube's Au oma ic Cap ions." In P oceedings o he
Fi s Wo kshop on E hics in Na u al Language P ocessing. Valencia, Spain: Associa ion o
Compu a ional Linguis ics, 2017. h ps://aclan hology.o g/W17-1606/
20
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
UNESCO and IRCAI. "Challenging Sys ema ic P ejudices: An In es iga ion in o Gende Bias in La ge
Language Models." Pa is: Uni ed Na ions Educa ional, Scien i ic and Cul u al O ganiza ion;
Ljubljana: In e na ional Resea ch Cen e on A i icial In elligence unde he auspices o UNESCO,
2024. h ps://unesdoc.unesco.o g/a k:/48223/p 0000388971
Sun, Luhang, Mian Wei, Yibing Sun, Yoo Ji Suh, Liwei Shen, and Sijia Yang. "Smiling Women Pi ching
Down: Audi ing Rep esen a ional and P esen a ional Gende Biases in Image-Gene a i e
AI." Jou nal o Compu e -Media ed Communica ion 29, no. 1 (2024):
zmad045. h ps://doi.o g/10.1093/jcmc/zmad045.
Uni ed Na ions De elopmen P og amme Se bia. "Rep oducing Inequali y: How AI Image Gene a o s
Show Biases Agains Women in STEM." Policy analysis. UNDP Se bia Blog. Belg ade: UNDP
Se bia, 2024. h ps://www.undp.o g/se bia/blog/ ep oducing-inequali y-how-ai-image-
gene a o s-show-biases-agains -women-s em.
Go ska, Anna M., and Da iusz Jemielniak. "The In isible Women: Unco e ing Gende Bias in AI-
Gene a ed Images o P o essionals." Feminis Media S udies 23, no. 8 (2023): 4370–
4375. h ps://doi.o g/10.1080/14680777.2023.2263659.
Melchio e, Alessand o Benede o, Na id Rekabsaz, Emilia Pa ada-Cabalei o, S e an B andl, Oleg
Leso a, and Ma kus Schedl. "In es iga ing Gende Fai ness o Recommenda ion Algo i hms in he
Music Domain." In o ma ion P ocessing & Managemen 58, no. 5 (2021):
102666. h ps://doi.o g/10.1016/j.ipm.2021.102666.
Lamb ech , Anja, and Ca he ine Tucke . "Algo i hmic Bias? An Empi ical S udy o Appa en Gende -
Based Disc imina ion in he Display o STEM Ca ee Ads." Managemen Science 65, no. 7 (2019):
2966–2981. h ps://doi.o g/10.1287/mnsc.2018.3093.
Eks and, Michael D., and Daniel Klu e . "Explo ing Au ho Gende in Book Ra ing and
Recommenda ion." Use Modeling and Use -Adap ed In e ac ion 31, no. 3 (2021): 377–
420. h ps://doi.o g/10.1007/s11257-020-09284-2.
Melchio e, Alessand o B., Na id Rekabsaz, Emilia Pa ada-Cabalei o, S e an B andl, Oleg Leso a, and
Ma kus Schedl. "In es iga ing Gende Fai ness o Recommenda ion Algo i hms in he Music
Domain." In o ma ion P ocessing & Managemen 58, no. 5 (2021):
102666. h ps://doi.o g/10.1016/j.ipm.2021.102666.
21
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
Smi h, Gene ie e. "Mindse s and Managemen : AI and Gende (In)Equi able Access o Finance."
Manusc ip accep ed o p esen a ion a ACM FAccT 2025. Upda ed July 17,
2025. h ps://a xi .o g/pd /2504.07312.
Ach a i, Ma gaux, Adil Salihu, Oli ie Mulle , Emmanuel Abbé, Ca ole Clai , Joëlle Schwa z, and
S ephane Fou nie . "Gende Bias in AI's Pe cep ion o Ca dio ascula Risk." Jou nal o Medical
In e ne Resea ch 26 (2024): e54242. h ps://doi.o g/10.2196/54242.
Da e, Nigam, and Sh ish i Kushwah. "The Feminiza ion o Gene a i e AI: Gende Delega ion Th ough
Names and Pe sonas." TechRxi p ep in . June 15,
2025. h ps://www. ech xi .o g/use s/928336/a icles/1299795- he- eminiza ion-o -gene a i e-
ai-gende -delega ion- h ough-names-and-pe sonas.
Buolamwini, Joy, and Timni Geb u. “Gende Shades: In e sec ional Accu acy Dispa i ies in Comme cial
Gende Classi ica ion.” In P oceedings o he 1s Con e ence on Fai ness, Accoun abili y and
T anspa ency, P oceedings o Machine Lea ning Resea ch 81:77–91.
2018. h ps://p oceedings.ml .p ess/ 81/buolamwini18a.h ml.
Te hö s , Philipp, Jan Niklas Kol , Nase Dame , Flo ian Ki chbuchne , and A jan Kuijpe . “Pos -
Compa ison Mi iga ion o Demog aphic Bias in Face Recogni ion Using Fai Sco e No maliza ion.”
P ep in , a Xi , Feb ua y 6, 2020. h ps://doi.o g/10.48550/a Xi .2002.03592.
Dha , P i h i aj, Joshua Gleason, Anike Roy, Ca los D. Cas illo, and Rama Chellappa. “PASS: P o ec ed
A ibu e Supp ession Sys em o Mi iga ing Bias in Face Recogni ion.” P ep in , a Xi , Augus 9,
2021. h ps://doi.o g/10.48550/a Xi .2108.03764
Thaku , Himanshu, A ishay Jain, P anee ha Vaddamanu, Paul Pu Liang, and Louis-Philippe Mo ency.
“Language Models Ge a Gende Makeo e : Mi iga ing Gende Bias wi h Few-Sho Da a
In e en ions.” P ep in , a Xi , June 7, 2023. h ps://doi.o g/10.48550/a Xi .2306.04597.
Jobin, Anna, Ma cello Ienca, and E y Vayena. “ The Global Landscape o AI E hics Guidelines.” Na u e
Machine In elligence 1, no. 2 (Sep embe 2019). h ps://doi.o g/10.1038/s42256-019-0088-2.
Boinodi is, Phaed a. “The Impo ance o AI Di e si y: D i ing T us wo hy AI.” IBM Consul ing. Accessed
Augus 23, 2025. h ps://www.ibm.com/ hink/insigh s/ai-di e si y-impo anceAu ho .
22
GEDIS - Gende Di e si y in In o ma ion Science:
Challenges in Highe Educa ion
P ojec Re e ence: 2024-1-ES01-KA220-HED-000246558
h ps://ub.edu/GEDIS
B ad o d, Nichol. “Why Di e si y in AI Makes Be e AI o All: The Case o Inclusi i y and Inno a ion.”
SHRM. Oc obe 7, 2024. h ps://www.sh m.o g/ opics- ools/ lagships/ai-hi/why-di e si y-in-ai-
makes-be e -ai- o -all-- he-case- o -inclus.
Suighi, Gio gio, and Rachel. “B eaking he Echo Chambe : Why Di e si y Is C ucial o AI’s Fu u e.” AIM
Resea ch Council. Janua y 8, 2025. h ps://council.aim esea ch.co/b eaking- he-echo-chambe -
why-di e si y-is-c ucial- o -ais- u u e/.
Hou, Tsung-Yu, Yu-Chia Tseng, and Chien Wen (Tina) Yuan. “Is This AI Sexis ? The E ec s o a Biased
AI’s An h opomo phic Appea ance and Explainabili y on Use s ’Bias Pe cep ions and
T us .” In e na ional Jou nal o In o ma ion Managemen 76
(2024). h ps://doi.o g/10.1016/j.ijin omg .2024.102775.
Eu opean Union. "AI Ac ." Shaping Eu ope's Digi al Fu u e. July 31, 2025. h ps://digi al-
s a egy.ec.eu opa.eu/en/policies/ egula o y- amewo k-ai.
C isan o, Juan Ca los, C is Benson Leu e io, Je my P enio, and Je e y Yong. “Regula ing AI in he
Financial Sec o : Recen De elopmen s and Main Challenges.” FSI Insigh s on policy
implemen a ion, no. 63. Decembe 2024. Bank o In e na ional
Se lemen s. h ps://www.bis.o g/ si/publ/insigh s63.pd .
an de Me we, Ma na, and Die e Veldsman. “AI Risk Managemen o HR: 3 Key Risks To Manage &
HR Ac ions To Take.” AIHR. Accessed Augus 23, 2025. h ps://www.aih .com/leading-h /ai- isk-
managemen /.
Soleimani, Mohammad, Ali In eza i, Julia A owsmi h, Daniel J. Pauleen, and Nes in Taskin. “Reducing
AI Bias in Rec ui men and Selec ion: An In eg a i e G ounded App oach.” The In e na ional
Jou nal o Human Resou ce
Managemen (2025). h ps://doi.o g/10.1080/09585192.2025.2480617.
