A Gendered Perspective on Artificial Intelligence

Parsheera, Smriti

doi:10.23919/itu-wt.2018.8597618

Cited by 11 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The final category considered papers proposing social solutions, with Figure 7 showing the emergent themes: 57% coded as improving algorithmic design and 33% as improving due process. The remainder fell under the improving diversity in datasets category, with the solution emerging as ensuring the data is representative (Avellan et al, 2020) and that gender bias is removed in cleaning up datasets (Parsheera, 2018). Within the algorithmic design theme, 4 main sub-themes emerged.…”

Section: Social Solutions To Algorithmic Gender Biasmentioning

confidence: 99%

A systematic review of socio-technical gender bias in AI algorithms

Hall

Ellis²

2023

OIR

View full text Add to dashboard Cite

PurposeGender bias in artificial intelligence (AI) should be solved as a priority before AI algorithms become ubiquitous, perpetuating and accentuating the bias. While the problem has been identified as an established research and policy agenda, a cohesive review of existing research specifically addressing gender bias from a socio-technical viewpoint is lacking. Thus, the purpose of this study is to determine the social causes and consequences of, and proposed solutions to, gender bias in AI algorithms.Design/methodology/approachA comprehensive systematic review followed established protocols to ensure accurate and verifiable identification of suitable articles. The process revealed 177 articles in the socio-technical framework, with 64 articles selected for in-depth analysis.FindingsMost previous research has focused on technical rather than social causes, consequences and solutions to AI bias. From a social perspective, gender bias in AI algorithms can be attributed equally to algorithmic design and training datasets. Social consequences are wide-ranging, with amplification of existing bias the most common at 28%. Social solutions were concentrated on algorithmic design, specifically improving diversity in AI development teams (30%), increasing awareness (23%), human-in-the-loop (23%) and integrating ethics into the design process (21%).Originality/valueThis systematic review is the first of its kind to focus on gender bias in AI algorithms from a social perspective within a socio-technical framework. Identification of key causes and consequences of bias and the breakdown of potential solutions provides direction for future research and policy within the growing field of AI ethics.Peer reviewThe peer review history for this article is available at https://publons.com/publon/10.1108/OIR-08-2021-0452

show abstract

Section: Social Solutions To Algorithmic Gender Biasmentioning

confidence: 99%

A systematic review of socio-technical gender bias in AI algorithms

Hall

Ellis²

2023

OIR

View full text Add to dashboard Cite

show abstract

“…Finally, the use of facial processing could also have an impact on other models that do not explicitly include gender as a parameter for decision making. Research has shown a gendered bias to be present in all sorts of algorithmic systems from recruitment to text embeddings and even hate speech detectors (Xia et al, 2020;De-Arteaga et al, 2019;Prates et al, 2019;Sap et al, 2019;Dixon et al, 2018;Parsheera, 2018). Many of these studies illustrate how algorithms can acquire a gender bias even when gender is not an explicit input.…”

Section: Gender Classificationmentioning

confidence: 99%

Cinderella's shoe won't fit Soundarya: An audit of facial processing tools on Indian faces

Jain¹,

Parsheera²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

The increasing adoption of facial processing systems in India is fraught with concerns of privacy, transparency, accountability, and missing procedural safeguards. At the same time, we also know very little about how these technologies perform on the diverse features, characteristics, and skin tones of India's 1.34 billion plus population. In this paper, we test the face detection and facial analysis functions of four commercial facial processing tools on a dataset of Indian faces. The tools display varying error rates in the face detection and gender and age classification functions. The gender classification error rate for Indian female faces is consistently higher compared to that of males -the highest female error rate being 14.68%. In some cases, this error rate is much higher than that shown by previous studies for females of other nationalities. Age classification errors are also high. Despite taking into account an acceptable error margin of plus or minus 10 years from a person's actual age, age prediction failures are in the range of 14.3% to 42.2%. These findings point to the limited accuracy of facial processing tools, particularly for certain demographic groups, and the need for more critical thinking before adopting such systems.

show abstract

“…Pese a que se puedan identificar referencias anteriores a ese momento, lo cierto es que es en 1955, en la Conferencia de Dartmouth(Parsheera, 2018), cuando se entiende adoptado terminológicamente el concepto de "inteligencia artificial", a manos del profesor John McCarthy.Desde entonces, hasta la actualidad, pese a la significativa evolución, especialmente en épocas recientes, la conceptualización no se ha asentado, destacando la ausencia de unanimidad en torno a este término (Hernández Giménez, 2019) y la proliferación de muy diversas definiciones sobre el término de IA (como examina Valls Prieto, 2021) que, si bien supone una riqueza conceptual, dificulta su aproximación, así como su regulación.No obstante, el estudio sobre esta sí se ha potenciado, por lo que, pese a la inexistencia de consenso en cuanto a la terminología, sí que se han realizado numerosos estudios en cuanto a su contenido, donde se acuerda que nos encontramos ante "sistemas que manifiestan un comportamiento inteligente"(Borges Blázquez, 2020, p. 55).En este sentido, pese a diversas categorizaciones existentes, destaca la distinción entre dos tipos de inteligencia artificial. Por un lado, la del tipo débil (o weak AI) -que sería equiparada con las capacidades humanas-por otro, la de tipo fuerte (o strong AI) -que superaría aquellas-.…”

unclassified

Discriminación algorítmica e ilusoria neutralidad. Actuales desafíos ante el fortalecimiento de tradicionales desigualdades en materia de violencia contra la mujer

López de Zubiría Díaz

2024

femeris

View full text Add to dashboard Cite

El empleo de herramientas de inteligencia artificial aplicadas a diversos ámbitos constituye una realidad patente, conforme su uso crece exponencialmente con carácter general en múltiples ámbitos de la vida. Al respecto, se presenta aquí un análisis sobre aquellas cuestiones más alarmantes en materia preventiva e intervencionista, con especial vinculación al ámbito de la violencia de género. Entre otras, las que derivan de encontrarnos ante un sistema significativamente opaco, pero sobre el que se confía cierta neutralidad que puede convertirse en un instrumento capaz de obstaculizar la igualdad, a través de la proliferación de tradicionales discriminaciones enmascaradas en ilusorias neutralidades promovidas por herramientas generadas a partir de una base sesgada y excluyente. Con ello, lejos de combatir este fenómeno delictivo, estaremos fomentando sus bases patriarcales desigualitarias a través de la creación de herramientas tecnológicas teóricamente neutrales pero discriminatorias en la práctica. Al respecto, sin ánimo de desarrollar una postura negacionista ante la evolución tecnológica en este texto, lo que se realiza aquí es un estudio sobre aquellos aspectos que deben atenderse para que dicho avance evite el crecimiento de discriminaciones y violencias simbólicas, base de la violencia contra las mujeres, enmascaradas en una suerte de inteligencia artificial, neutral, que realmente esconda marcados sesgos de género

show abstract

A Gendered Perspective on Artificial Intelligence

Cited by 11 publications

References 7 publications

A systematic review of socio-technical gender bias in AI algorithms

A systematic review of socio-technical gender bias in AI algorithms

Cinderella's shoe won't fit Soundarya: An audit of facial processing tools on Indian faces

Discriminación algorítmica e ilusoria neutralidad. Actuales desafíos ante el fortalecimiento de tradicionales desigualdades en materia de violencia contra la mujer

Contact Info

Product

Resources

About