Why closing the gender gap in AI requires immediate action?
Every
area of the economy has experienced extraordinary levels of digitalization over
the past five years, spurred on by the pandemic. Private investment in the
industry totaled $93.5 billion in 2021 alone, more than quadruple the amount in
2020. The technology driving the digital transition, artificial intelligence
(AI), leaves women behind at every stage of its life cycle.
The
gender gap in AI is a self-sustaining problem. And it runs the risk of leaving
us with a technological and economic system in which women are vastly
underrepresented.
The lack of women in AI
Women
are less active online than men. The epidemic hastened the uptake of mobile
internet, but according to GSMA data, women are 16% less likely than males to
use mobile internet in low- and middle-income countries. Additionally, the
gender gap in industry 4.0 has made it much harder for women to break into the
tech industry.
With
the percentage of female PhDs in computer science and artificial intelligence
(AI) remaining at 20% for the previous ten years, this enormous disparity is a
problem that has not been resolved.
These
are serious issues in and of themselves, but they also have repercussions on
other aspects of the AI life cycle, particularly in terms of innovation and
advancement. The percentage of men who graduate in information and
communication technologies (ICT) is 400% higher than that of women (8.2% versus
1.7%), according to recent study by the World Economic Forum.
Only
one-quarter of all tech occupations are held by women, and men continue to hold
the majority of technical and leadership positions in the industry.
Particularly, only 22% of AI professionals worldwide are women. Only 18% of
writers at the top AI conferences are female, while only 13.83% of AI papers
are written by women. In 2019, only 2% of venture financing went to start-ups
run by women.
These
figures violate fundamental ideas of inclusion and diversity. But the industry
as a whole suffers from the lack of female participation as well, as it becomes
more effective the more gender varied it is.
Threatened
to intensify this trend even further is COVID-19. Due to the fact that women in
tech are twice as likely as males to have lost their jobs and that 42% of them
claim to have handled the majority of domestic duties during the epidemic, the
inequity women suffer at work is exacerbated by the imbalance they feel at
home.
Gender
biases in data sets and AI algorithm products, as well as the lack of gender
diversity in the workforce, gender inequities in STEM education, and the
inability to address the unequal distribution of power and leadership in the AI
sector, are all grave concerns. These systems reproduce gender prejudice in
ways that can widen the existing gender gap by propagating and strengthening
negative gender preconceptions.
Taking systemic inequalities into account
Addressing
the inequality at every stage of the life cycle is the main task in this
situation. To effectively and efficiently fight these problems, we must
legislate equality and advance diversity.
In
order to promote women and girls' full participation in the digital sector, it
is imperative that concrete policy measures be put in place. At the same time,
it is important to address ingrained stereotypes and social norms that
persistently contribute to discrimination and even violence against women.
Beyond national laws, we require a universal standard—a normative tool—to
direct the creation and application of AI, ensuring that they serve the
interests of all people, particularly women and girls.
This
problem has been accepted by UNESCO.
The
Recommendation on the Ethics of Artificial Intelligence, the first global
standard-setting document for AI, was created at the request of its members and
adopted by all 193 Member States in November 2021.
The
recommendation establishes the foundation for a digital transition that
supports environmental sustainability, gender equality, human rights, and human
dignity. It accomplishes this by using a related set of moral standards to
direct the ethical creation and application of artificial intelligence.
In
an important way, the recommendation urges governments to support gender
equality, the idea of "ensuring diversity and inclusiveness," and the
rule of "fairness and non-discrimination," laying the foundation for
an ethical gender-equality imperative in the digital age.
The
advice includes the following specific steps to address this issue:
- · Allocating
monies from the public budgets tied to sponsoring gender-related programs.
- · Ensuring
that a gender action plan is included in national digital policies.
- · Addressing
the issue of the pay and opportunity inequities at work, which have long been a
significant barrier to women's economic empowerment.
- · Promoting
female entrepreneurship, involvement, and engagement throughout the whole AI
life cycle.
- · Investing
in programmes that are specifically designed to expand chances for girls and
women to participate in STEM and ICT fields.
- · Eliminating
gender stereotypes and making sure that prejudices are not reflected in AI
systems.
In
order to guarantee that AI is used to advance the rights of all people,
especially women and girls, UNESCO is also creating two ground-breaking tools:
Readiness Assessment Methodology and Ethical Impact Assessment.
We
have a decision to make today between progress and digitally reproducing the
injustices and discrimination of the physical world. The gender gap in digital
technology has always existed and will continue to persist unless we change the
gendered societal norms, despite the market's increasing availability of this
technology.
We
must act now to catch up with the quickly growing industry, especially with the
tools provided by UNESCO to close the gender gap in AI. As stated in the Global
Gender Gap Report 2022, we cannot wait another 132 years to close the gender
gap globally.