How does AI impact gender equality?
Intellectual Forum Research Associate Clementine Collett (2017) has worked with Dr Sarah Dillon to publish a new report, ‘AI and Gender: Four Proposals for Future Research’. The report focuses on the issues which need to be untangled and subsequently addressed when it comes to the relationship between AI and gender equality. Here, Clementine tells us about her research.
The development of AI does, and will, impact on issues surrounding gender equality. In parallel, existing inequalities and injustices influence the way in which AI is developed.
Our report focuses on four of the most significant challenges to gender equality presented by recent developments in AI. In tandem, it suggests four academic research proposals which would effectively tackle these issues. These proposals aim to facilitate practical action on issues of injustice resulting from the design, purpose, and use of AI.
1. Bridging Gender Theory and AI Practice
Technological design often captures and reproduces controlling and restrictive conceptions of gender and race which are then repetitively reinforced. We propose research which utilises gender theory, including trans, queer and feminist theory, alongside multilateral conversations with international stakeholders, to explore fundamental barriers to equality embedded in these technologies.
2. Law and Policy
There is a risk that economic prosperity and political power will play an underlying role in shaping laws and policies concerning AI, at the expense of other, more equalising, motivations. We suggest there is a need for research which analyses existing and emerging legislation and policy related to AI which will have an impact on gender equality. Specifically, surrounding data and privacy, technological design, and labour.
3. Biased Datasets
Datasets are often unrepresentative of the public demographic, and there is a high level of data deprivation when it comes to capturing vulnerable groups. Context-specific, gender-specific guidelines for best practice regarding data collection, handling and trade-offs need to be established.
4. Diversity in the AI Workforce
Currently, there is a significant gender disparity in the AI workforce. Research conducted by Element AI has discovered that around 12% of AI researchers are women. And last year, PwC surveyed 2000 A-level students in the UK, finding that only 27% of female students said they would consider a career in technology, in contrast to 61% of males. Diversification of the AI workforce will be vital in order to design and implement technology which is equitable. There is a need for research which explores the factors that impact diversity in STEM education and in the AI workforce, and which considers how to create a sustainable culture of diversity in institutions and in the workplace.
So, where did this report and this initiative come from?
Back in February 2019, I helped to convene the Leverhulme Centre for the Future of Intelligence in Cambridge workshop on ‘AI and Gender’. This was convened by the Ada Lovelace Institute and supported by PwC. Over the day, there were 17 brilliant 10-minute talks. All of them gave unique and fascinating insights into the topic of AI and Gender, often introducing fresh-off-the-press, cutting-edge research.
The workshop was trans-disciplinary and trans-sectoral. It gathered together scholars from several fields, including computer science, history, philosophy of science, law, politics, sociology, literature and gender studies. In addition, it brought together researchers and practitioners from industry and research centres outside of academia, as well as key figures from UK AI governance and policy.
Towards the end of the conference we set the participants a challenge. We labelled this the ‘collective intelligence activity’. After dividing the participants into four groups, we asked them for at least three recommendations for new areas of research on AI and Gender. This report develops and augments the ideas shared during the collective intelligence activity with content from the presentations and discussions from the day, alongside a broad survey of relevant literature and research.
The media is constantly bombarding us with images of how rapid technological development will change the world in which we live, often leaving us feeling that there’s not much we can do about it. Now is the moment to address the existing and potential issues of injustice which stem from the deployment of AI technology, and to ensure that we play a role in shaping our own future.
The opinions expressed in this piece are those of the author.