78
%
of women
perceive gender bias
Impact of AI across eight diversity dimensions
This first report by the Observatory on AI and Diversity analyses the social, functional, and symbolic impact of artificial intelligence on people belonging to eight diversity dimensions:
GENDER AND SEXUAL ORIENTATION
WOMEN
RELIGIOUS DIVERSITY
DISABILITY
NEURODIVERSITY
AGE
ETHNIC
DIVERSITY
DIGITAL
LITERACY
🧑🦱 Sample: 157 people
🧑🎓 Validation: Expert Committee
🧠 Objective: Inclusive AI
Focus on Urban Mobility
The report also examines how these biases specifically affect urban mobility, revealing access barriers, algorithmic exclusion, and opportunities for inclusive redesign.
Study Objectives
Detect
algorithmic
bias
Assess
accessibility
and inclusion
Propose
improvements
towards ethical AI
Key Findings
60
%
of people with disabilities
fear automated discrimination
75
%
believe AI is not accessible for older people
only
35
%
have seen non-binary options in interfaces
46
%
received stereotyped recommendations
59
%
detected visual bias
in AI-generated images
93
%
use AI,
but the majority do not
trust it
trust it
AI and Urban Mobility
56
%
have used
shared
mobility
32
mentions of lack
of physical
accessibility
35
mentions
of technological
barriers
19
mentions
of technological
barriers
Key Conclusions
AI IS NOT
NEUTRAL
NEUTRAL
REPRESENTATION
MATTERS
MATTERS
THE DIGITAL DIVIDE IS MULTIDIMENSIONAL
ACTIVE PARTICIPATION
OF ALL GROUPS
IS ESSENTIAL
OF ALL GROUPS
IS ESSENTIAL
INCLUSIVE DESIGN IS NOT OPTIONAL