Attesting Biases and Discrimination using Language Semantics

Xavier Ferrer Aran; Jose M. Such; Natalia Criado Pacheco

Attesting Biases and Discrimination using Language Semantics

Xavier Ferrer Aran, Jose M. Such, Natalia Criado Pacheco

Research output: Contribution to journal › Conference paper › peer-review

112 Downloads (Pure)

Abstract

AI agents are increasingly deployed and used to make auto-
mated decisions that affect our lives on a daily basis. It is imperative to
ensure that these systems embed ethical principles and respect human
values. We focus on how we can attest whether AI agents treat users fairly
without discriminating against particular individuals or groups through
biases in language. In particular, we discuss human unconscious biases,
how they are embedded in language, and how AI systems inherit those bi-
ases by learning from and processing human language. Then, we outline a
roadmap for future research to better understand and attest problematic
AI biases derived from language.

Original language	English
Pages (from-to)	1
Number of pages	8
Journal	Autonomous Agents and Multi-Agent Systems
Publication status	Accepted/In press - 17 Mar 2019

Keywords

digital discrimination
bias
ethics
agents
NLP

Access to Document

Attesting Biases and Discrimination_FERRER_Accepted17March2019_GREEN AAMAccepted author manuscript, 194 KB

DADD: Discovering and Attesting Digital Discrimination
Such, J. (Primary Investigator), Tasioulas, J. (Co-Investigator), Nelken, D. (Co-Investigator), Viganò, L. (Co-Investigator), Criado Pacheco, N. (Co-Investigator), Hedges, M. (Co-Investigator) & Coté, M. (Co-Investigator)
EPSRC Engineering and Physical Sciences Research Council
1/08/2018 → 31/05/2022
Project: Research

Cite this

@article{50151f28e06546fea1bbfa1e3b7b33fe,

title = "Attesting Biases and Discrimination using Language Semantics",

abstract = "AI agents are increasingly deployed and used to make auto-mated decisions that affect our lives on a daily basis. It is imperative toensure that these systems embed ethical principles and respect humanvalues. We focus on how we can attest whether AI agents treat users fairlywithout discriminating against particular individuals or groups throughbiases in language. In particular, we discuss human unconscious biases,how they are embedded in language, and how AI systems inherit those bi-ases by learning from and processing human language. Then, we outline aroadmap for future research to better understand and attest problematicAI biases derived from language.",

keywords = "digital discrimination, bias, ethics, agents, NLP",

author = "{Ferrer Aran}, Xavier and Such, {Jose M.} and {Criado Pacheco}, Natalia",

year = "2019",

month = mar,

day = "17",

language = "English",

pages = "1",

journal = "Autonomous Agents and Multi-Agent Systems",

issn = "1387-2532",

publisher = "Springer Netherlands",

}

TY - JOUR

T1 - Attesting Biases and Discrimination using Language Semantics

AU - Ferrer Aran, Xavier

AU - Such, Jose M.

AU - Criado Pacheco, Natalia

PY - 2019/3/17

Y1 - 2019/3/17

N2 - AI agents are increasingly deployed and used to make auto-mated decisions that affect our lives on a daily basis. It is imperative toensure that these systems embed ethical principles and respect humanvalues. We focus on how we can attest whether AI agents treat users fairlywithout discriminating against particular individuals or groups throughbiases in language. In particular, we discuss human unconscious biases,how they are embedded in language, and how AI systems inherit those bi-ases by learning from and processing human language. Then, we outline aroadmap for future research to better understand and attest problematicAI biases derived from language.

AB - AI agents are increasingly deployed and used to make auto-mated decisions that affect our lives on a daily basis. It is imperative toensure that these systems embed ethical principles and respect humanvalues. We focus on how we can attest whether AI agents treat users fairlywithout discriminating against particular individuals or groups throughbiases in language. In particular, we discuss human unconscious biases,how they are embedded in language, and how AI systems inherit those bi-ases by learning from and processing human language. Then, we outline aroadmap for future research to better understand and attest problematicAI biases derived from language.

KW - digital discrimination

KW - bias

KW - ethics

KW - agents

KW - NLP

M3 - Conference paper

SN - 1387-2532

SP - 1

JO - Autonomous Agents and Multi-Agent Systems

JF - Autonomous Agents and Multi-Agent Systems

ER -

Attesting Biases and Discrimination using Language Semantics

Abstract

Keywords

Access to Document

Fingerprint

Projects

DADD: Discovering and Attesting Digital Discrimination

Cite this