Main Content

Artificial intelligence “doubled-edged sword,” for addressing gender, racial bias, report shows

September 23, 2021

Toronto – From social media, to health diagnostics, policing, and assessing job applications, artificial intelligence has spread rapidly into many facets of daily life.  

It brings efficiency and the capacity to do good. AI technologies can overcome human error as well as biases that can taint human decision-making, especially those with negative impacts on other people.  

But care and attention is urgently needed by businesses and governments to curb the technologies’ equal potential to reinforce gender and racial inequities, says a new report synthesizing related research to date.  

“AI has a ton of power to create outcomes that are very helpful to people,” says Carmina Ravanera, the report’s co-author and a research associate at the Institute for Gender and Economy (GATE) at the University of Toronto’s Rotman School of Management.  

“But, we can’t think of technology as separate from the issues going on in society,” she says. “It cannot be assumed that a new technology is objective and free from taking on biases that have existed for a very long time.”  

AI describes machine-based technologies that “learn” from data they receive, adapting and refining their subsequent responses within a set of programmed rules and restrictions called algorithms.  

Those responses are only as neutral as the data and algorithms the technology is trained on, the report shows. A British medical school rejected women and applicants with non-European-sounding names because the algorithm used by its computer screening tool was based on data where such applicants were significantly underrepresented. In another example, an AI system for detecting cancerous skin lesions was found to be less likely to pick up cancers in dark-skinned people because the system had been developed from a database comprised of mostly light-skinned populations.  

Bias and discrimination can also be perpetuated through implementation, such as facial analysis software used to predict who will be more prone to criminality based on facial features.  The introduction of AI into the workplace, meanwhile, may lead to higher job loss for women and members of racialized groups because of their strong representation in occupations vulnerable to automation.  

The report recommends that potentially detrimental applications of AI can be avoided by organizations embedding social equity considerations from an AI project’s start and increasing representation of gender diverse and racialized groups within project teams.  

Governments, meanwhile, need to get going on policy and regulations that will set standards and accountabilities for these new technologies, says the report, pointing to recent examples such as the United States’ 2019 Algorithmic Accountability Act and a proposed Artificial Intelligence Act by the European Union this year.  

It’s a smart thing to do, says Ms. Ravanera, because questions about AI fairness are getting louder and more frequent in public discourse, increasing the pressure to act.  

“There are always ways in which governments and organizations creating and using AI can take a pause and say, ‘Hey, we shouldn’t do this,’” she says. “It’s not too late. You can always make changes.”  

The report, “An Equity Lens on Artificial Intelligence,” was co-authored with Sarah Kaplan, Distinguished Professor of Gender and the Economy at the Rotman School and director of Institute for Gender and the Economy.   

The report is available online at  

The Institute for Gender and the Economy (GATE) at the University of Toronto’s Rotman School of Management promotes an understanding of gender inequalities and how they can be remedied—by people of all genders—in the world of business and, more broadly, in the economy. At GATE, we are changing the conversation on gender equality by: using rigorous research to investigate the hidden mechanisms that propagate gender equality; funding, translating, and disseminating innovative, academic research; and engaging executives, policy makers, and students to create new solutions for achieving equality, advancing careers, and creating economic prosperity. Visit for more information.   

Bringing together high-impact faculty research and thought leadership on one searchable platform, the new Rotman Insights Hub offers articles, podcasts, opinions, books and videos representing the latest in management thinking and providing insights into the key issues facing business and society.  


The Rotman School of Management is part of the University of Toronto, a global centre of research and teaching excellence at the heart of Canada’s commercial capital. Rotman is a catalyst for transformative learning, insights and public engagement, bringing together diverse views and initiatives around a defining purpose: to create value for business and society. For more information, visit  



For more information:  

Ken McGuffin
Manager, Media Relations
Rotman School of Management
University of Toronto