The role of AI, technology and education in gender equality

For two weeks in March, I had the pleasure of being chosen from over 3,000 applicants as a UN Women UK Official Delegate for the 67th Meeting of The UN Commission for the Status of Women (CSW67). This amazing global meeting of independent advocates, representatives of government and groups was focused on the primary theme of ‘Innovation and technological change, and education in the digital age for achieving gender equality and the empowerment of all women and girls.’

UN Women UK is the United Kingdom’s national committee of UN Women, a subset of the United Nations group dedicated to gender equality and women's empowerment. UN Women UK works towards promoting gender equality and women's rights in the UK and globally through various initiatives, campaigns and partnerships, advocating for policy change, raising awareness of gender-based discrimination and violence and promoting women's economic empowerment.

A small but mighty charity, they work towards achieving gender equality in all spheres of life including leadership, education, health and political participation. Additionally, UN Women UK work to engage men as allies in empowering women through initiatives such as #heforshe.

Throughout the two weeks, we were treated to a range of events covering different talking points in technology, including AI and data. The core theme emphasised the need for innovation and education in areas such as data literacy to be leveraged to promote gender equality, empower women and girls, and address gender-based violence.

The intersection of tech, gender and equality

The intersections of technology, gender and equality can be complex, especially in relation to AI. A core theme of CSW67 was the lack of diversity in teams developing AI systems; one of the main challenges for the technology is its potential to perpetuate existing biases and inequalities. There is a growing concern that AI systems have been proven to be biased against women, minorities, and other marginalised groups.

For example, image generators or large language models (LLM) are trained on datasets filled with these biases. When algorithms are taught on data sets that contain biases, they are likely to reproduce them in their output. This is why it is essential to ensure that the data models draw from is representative of a wider subset of society to avoid amplifying existing problems.

Examples were given in online talks by groups such as Feminist Task Force. The group’s speakers particularly highlighted the issues faced by asylum seekers trying to access the US government migration app, ‘CBP One’, particularly in Haiti and at the US-Mexico border. One of the top issues faced by families and individuals using the app is errors in recognising faces with darker skin tones. Repetitive failure to accept entries and identify users has caused major delays in registering and processing applications, with calls now being made by legal and advocacy groups for investigations into failures.

Speaking at “A Gender Equal World with Technologies, Digitalisation and AI”, Vera Jourová of the European Commission focused on gender stereotypes, AI algorithms and the tremendous potential that technology and digitalisation has for our lives. She discussed the importance of preventing new technology from perpetuating gender stereotypes and bias, emphasising that work in this area should not undermine gender equality or the democratic model.

The European Commission want the roadmap for a global digital compact to be gender transformative, and Jourová mentioned that the EU promotes a human-centred approach to the digital transition. There are at least two proposals, according to Jourová, which are being suggested to address the gender digital divide by EU co-legislators. One of these, The Artificial Intelligence Act, covers reinforcement of biases and how they may be counteracted.

A digital education action plan, which focuses on the digital readiness of education and training systems and encourages girls to be educated in STEM, is also currently in progress and running up to 2027. The strategy includes targets to address the digital divide, including aiming for 80% of the EU population to have basic digital skills by 2030, with 20 million becoming ICT specialists — and this number should be gender balanced.

Fighting bias in AI

To address this subject and other technology-related concerns around gender equality, UN Women UK organised a special in-person event at The Roundhouse in London, attended by a select group of 250 participants chosen from the nearly 3,000 UK delegates who took part this year. The event facilitated in-depth discussions on leveraging technology to promote gender equality without perpetuating existing biases.

For you

Be part of something bigger, join BCS, The Chartered Institute for IT.

The focal point of the AI team’s discussion was how to train algorithms, which often rely on biased human databases, to generate unbiased outputs. Approaches to addressing bias in AI models vary, but the ideas included using more diverse and representative datasets to train algorithms. This would ensure that the data used to train the algorithm was reflective of the population it serves. For example, if an algorithm is designed to detect illnesses such as cancer, it is crucial to use data sets that are representative of all groups, including different races, ages, and genders. This approach could ensure that the algorithm is effective for everyone and does not perpetuate existing biases.

There was also general enthusiasm for regulation to ensure organisations follow ethical guidelines, which ought to be designed to ensure that algorithms are developed and used in a transparent manner. This is essential when considering the implications for data used to train algorithms; for instance, companies that use algorithms to predict consumer behaviour need to ensure that they are not engaging in unethical practices such as discrimination.

The importance of education

CSW67 also highlighted the importance of education and technology in promoting gender equality. Education plays a critical role in shaping attitudes and perceptions about gender roles and expectations, and it is essential to ensure that education systems promote gender equality and challenge gender stereotypes. Additionally, the use of technology in education, such as e-learning and online courses, can help to promote gender equality by providing access to education for women and girls in areas where traditional education is not available.

Technology can also help empower women by providing them with access to information and resources; mobile apps can provide women with access to health information and services, financial resources, and legal support. These technologies can help women make informed decisions about their lives, improve their economic status, and reduce their vulnerability to violence and discrimination.

In conclusion, the 67th meeting of the United Nations Commission for the Status of Women highlighted the critical role of AI, technology and education in achieving gender equality and the empowerment of all women and girls. While technology has the potential to be good for all society, it is vital to ensure that decision-makers, designers, and the data they draw from are representative of all groups to avoid perpetuating gender stereotypes and existing biases.