Impact and Perspectives of Women and Artificial Intelligence

Impact and Perspectives of Women and Artificial Intelligence Presented By: loana Stupariu, LL.M Date: 8 March 2024 Protecting Women's Rights - As a particularly vulnerable group, women have special status and protection within the United Nations and regional human rights systems • The Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) - 1979 • The Declaration on the Elimination of Violence Against Women (DEVAW) 1993 • - The Istanbul Convention - 2014 Rights covered by the aforementioned: • Equality and non-discrimination • Right to life, liberty, and security • Employment rights • Participation in political and public life • Health rights • Equality in marriage and family relations • Education rights • Economic and social benefits Al Unveiled: Defining Artificial Intelligence and Its Potential to Transform Women's Rights - Al, or Artificial Intelligence, refers to the capability of computers or robots controlled by computers to - perform tasks that typically require human intelligence. - In the Al Act, aligning with internationally recognized criteria, following OECD guidelines, an Al system is described as following: - "a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments." • Potential for advancing gender equality by automating tasks, providing educational and career opportunities, and addressing biases. • Might perpetuate gender stereotypes and increase unemployment in sectors with high female employment. Al Impact on Women's Rights at Work - 79% of working women are in occupations vulnerable to Al disruption. - Generative Al's impact on jobs shows a higher risk for women than men in the U.S. workforce, with 58.87 million women in high-risk occupations versus 48.62 million men. - 21% more women than men are exposed to Al automation, primarily due to the higher percentage of women in white-collar roles susceptible to automation. - Sectors like healthcare, education, and administrative support, predominantly staffed by women, are at high risk. - Generative Al could automate tasks in these fields, impacting job security for women. - Despite risks, Al offers opportunities for job creation and enhancing job satisfaction by automating repetitive tasks. - Women are encouraged to adapt by learning to use Al effectively. Al Poses Disproportionate Risks to Women - Al-related crimes and biases exacerbate inequalities and introduce new harassment and discrimination forms against women. • Online Harassment and Cyberbullying • Employment Discrimination • Deepfakes • Manipulation of Social Media Algorithms - • Al-Enabled Domestic Violence • Stalker ware and Spyware • Algorithmic Invasions of Privacy • Al-driven phishing Attacks • Health Data Exploitation Deepfakes have become a sophisticated form of gender-based violence, predominantly targeting women in explicit videos. - 96% of deepfakes are non-consensual sexual videos, with 99% targeting women, spotlighting gender-based violence. The Digital Gender Divide and the role of Al - Al can reinforce stereotypes about women, affecting their chances to learn new skills or pursue careers in science, technology, engineering, and mathematics (STEM). • Women have less access to technology! (UNESCO, OECD, IDB (2022). The Effects of Al on the Working Lives of Women) • Internet access disparities: Women in Africa (20.2%) vs. Europe (80.1%) • Limited access to education, digital skills, and job opportunities. • Low representation and leadership in Al fields. Only 12 percent of Al researchers are women! • Women make up 25% of the STEM workforce but only 9% of leaders in these fields. • According to a 2021 Stanford study, just 16% of tenure-track faculty focused on Al are women. • Governments, the private sector, and other actors must make efforts to ensure women are not left behind in the digital economy. Gender Bias and Artificial Intelligence - AI BIAS-a phenomenon that occurs when an algorithm produces results that are systemically prejudiced due to erroneous assumptions in the machine learning process. - Al biases can lead to unequal treatment for women and the reinforcement of gender stereotypes. - "Hey Siri, go make me a sandwich." - Why are so many Al assistants female? - "The use of female names is another way to feminize and domesticate Al. This form of gendering comes with certain expectations and stereotypes, and has the potential to perpetuate existing gender inequalities and biases in society." - Algorithmic transparency is a must if we want to improve women's digital and Al skills, and to ensure gender-aware algorithms and related policies. Underrepresentation of Women in Tech - Women are significantly underrepresented in the global Al workforce despite their contributions. The NYT published in December 2023 an article with the 12 most influential people in Al. How many women were featured? Case Study: Women Receiving Less High Paying Job Ads - A Carnegie Mellon University (CMU) study reveals gender discrimination in online ad targeting, showing high-paying job ads more frequently to males. - When Google presumed users to be male job seekers, they were much more likely to be shown ads for high-paying executive jobs. Google showed the ads 1,852 times to the male group - but just 318 times to the female group. - Who was responsible in this case? Was it the preference of advertisers? Or was it the unintended consequence of machine learning algorithms that drive online recommendation engines? - Though the tool developed in the study can identify discrepancies in online algorithms, it can't explain why they occur without a look inside the black box. A More Equitable Digital Future for All - We need to ensure a digital future that is equitable and accessible to everyone, irrespective of gender, race, or socio-economic status. Achieving this requires a collective effort from governments, industry, and civil society to foster an environment where technology amplifies diversity, equity, and inclusion. • Importance of public engagement and awareness: in order to combat bias, we need to be aware of it. • Laws and regulations need to evolve at the same rate as technology does. • Diverse data sets. Women participation in Al research and development. Ethical guidelines to prioritize non-discrimination. Support for women-led Al startups. Equitable access to Al education and career opportunities. • Prioritizing gender equality and inclusivity at every stage of Al development and deployment. • Regulatory frameworks must enforce transparency in Al algorithms.