The gen AI gender gap

The gen AI gender gap By Iñaki Aldasoro, Olivier Armantier, Sebastian Doerr,Leonardo Gambacorta and Tommaso Oliviero Monetary and Economic Department July 2024 JEL classification: C8, D8 Keywords: artificial intelligence, privacy, gender, gen AI BIS Working Papers are written by members of the Monetary and Economic Department of the Bank for International Settlements, and from time to time by other economists, and are published by the Bank. The papers are on subjects of topical interest and are technical in character. The views expressed in them are those of their authors and not necessarily the views of the BIS. This publication is available on the BIS website (www.bis.org). The gen AI gender gap* ˜ Inaki Aldasoro Olivier Armantier Sebastian Doerr Leonardo Gambacorta Tommaso Oliviero Abstract Generative artificial intelligence (gen AI) is expected to increase produc- tivity. But if unequally adopted across demographic groups, its prolifer- ation risks exacerbating disparities in pay and job opportunities, leading to greater inequality. To investigate the use of gen AI and its driverswe draw on a representative survey of U.S. household heads from theSurvey of Consumer Expectations. We find a significant “gen AI gen- der gap”: while 50% of men already use gen AI, only 37% of women do.Demographic characteristics explain only a small share of this gap, while respondents’ self-assessed knowledge about gen AI emerges as themost important factor, explaining three-quarters of the gap. Gender dif-ferences in privacy concerns and trust when using gen AI tools, as wellas perceived economic risks and benefits, account for the remainder.We conclude by discussing implications for policy to foster equitable gen AI adoption. JEL classification: C8, D8. Keywords: artificial intelligence, privacy, gender, gen AI. 1 Introduction Generative artificial intelligence (gen AI) holds the potential to boost eco- nomic activity. Evidence suggests it makes workers more productive, espe- cially in occupations that require cognitive abilities (Brynjolfsson et al.,2023;Noy and Zhang,2023;Peng et al.,2023), and spurs firm growth and innova- tion (Babina et al.,2024). Gen AI is thereby poised to have profound effects on aggregate output and wages (Baily et al.,2023;Aldasoro et al.,2024). A key concern, however, is that increasing AI adoption will lead to greater inequality (Cazzaniga et al.,2024), especially if unequally adopted acrossde- mographic groups. For example, previous work has shown stark differences in the use of financial technology (fintech) between men and women – leading to a “fintech gender gap” (Chen et al.,2023). If there are similardisparities in gen AI usage, it could exacerbate existing differences in pay and job opportu- nities. To assess who will benefit from gen AI and how it willshape inequality, it is thus crucial to understand who uses it and why (not). This paper investigates gender differences in the use of gen AI and their drivers, based on a representative survey of U.S. consumers. It draws on ques- tions that were added to the Federal Reserve Bank of New York’s Surveyof Consumer Expectations (SCE), fielded in February 2024. Our key finding is the presence of an economically and statistically signif- icant “gen AI gender gap”.Figure 1 shows that women are significantly less likely than men to use gen AI. On average, 50% of men report having usedgen AI over the previous twelve months. The respective number for womenis 37% (panel (a)). A significant gender gap is also present among frequentusers of gen AI, i.e., those that have used gen AI weekly over the previous twelve months (panel (b)). What explains the gender gap in the use of gen AI technology? We find thatit is not driven by demographic characteristics such as income, educa- tion,age, or race. Instead, respondents’ knowledge about gen AI emerges as themost important driver of the gap, explaining almost three-quarters. This result echoes findings of a gender gap in the use of technology more broadly (Scheerder et al.,2017;Lythreatis et al.,2022). The remainder of the gap is ex- plained by gender differences in attitudes towards privacy and trustin coun- terparties, consistent with previous findings that women are generally more concerned about the negative consequences of sharing data (Armantier et al.,2021,2024;Aldasoro et al.,2024;Prince and Wallsten,2022;Ta ng,2024); as well as perceived opportunities and risks from gen AI for employment. These findings suggest that gen AI could amplify the gender pay gap. To address this issue, privacy regulations as well as policies that promote AI- related knowledge and skills could be put in place. These could help to ensureequal opportunity for everyone to benefit from the capabilities of gen AI. 2 Data, Empirical Strategy and Results The Survey of Consumer Expectations.We investigate the use of gen AI with data from the Survey of Consumer Expectations. The SCE is a high- quality monthly,