Building Partnerships to Mitigate Bias in AI
Artificial intelligence (AI) is increasingly employed to make decisions affecting most aspects of our lives, particularly as COVID-19 accelerates the global digital transformation. Using technical systems such as machine learning, AI is becoming progressively more involved in deciding who receives an interview for a job, whether someone will be offered credit, which products are advertised to which consumers, as well as how government services and resources are allocated, which neighborhoods are targeted as “high risk” for crime, and more.
The use of AI in decision-making can reduce human subjectivity, but it can also do the exact opposite when AI systems are powered by biased datasets and algorithms, producing discriminatory outcomes at scale and posing immense risks to businesses and societies. Harnessing the transformative potential of AI requires addressing these biases. As developers, users and managers of AI systems, businesses play a central role in leading the charge while business leaders’ decisions are of historical importance.
In this sense, the Center for Equity, Gender & Leadership of the University of Berkeley in California organized an event on 16 July 2020 to launch the Mitigating Bias in Artificial Intelligence (AI): An Equity Fluent Leadership Playbook, which helps business leaders address bias in AI to unlock value responsibly and equitably.
Moderated by Genevieve Smith, Associate Director of the Center, this event gathered stakeholders and experts from the private and public sectors and academia. Ms Saniye Gülser Corat, UNESCO’s Director for Gender Equality participated as one of the four panelists, alongside Mr Donald Martin Jr., Sr. Technical Program Manager & Social Impact Technology Strategist at Google; Ms Deborah Raji, Tech Fellow at the AI Now Institute; and Ms Katia Walsh, Chief Strategy & AI Officer at Levi Strauss & Co. In her intervention, Ms Corat talked about how UNESCO is leveraging its convening power in order to build partnerships that help address gender biases in AI and machine learning.
Fostering Conversation on the Gendering of AI
UNESCO became the first United Nations agency to work on the gendering of Artificial Intelligence with the launch of a publication in March 2019 entitled I’d Blush if I Could: closing gender divides in digital skills through education, which features recommendations on actions to overcome global gender gaps in digital skills, with a special examination of the impact of gender prejudice coded into some of the most prevalent artificial intelligence (AI) applications such as digital voice assistants.
The publication locates this prejudice in the gender imbalance of technical teams developing frontier technologies, which often leads to the replication of stereotypes into the machines. Since its launch, I’d Blush If I could has helped spark a global conversation, particularly with the private sector, on the gendering of AI technology and the importance of education to develop the digital skills of women and girls.
To promote this global conversation, UNESCO participated at the Web Summit 2019, the world’s largest gathering of entrepreneurs, which brought together more than 70.000 participants in Lisbon in November 2019. During a fireside chat moderated by Ms Esther Paniagua from El País, Ms Saniye Gülser Corat, UNESCO’s Director for Gender Equality, explained that at a moment when every sector is becoming a technology sector, these gender equality concerns are becoming very alarming. For her, “to build a more equal future, women need a seat at the table. They need to be involved in designing tech, so that they can play a role in steering them”.
In 2020, UNESCO decided to take this conversation further by looking at what concrete actions the private and public sector should take in order to ensure that AI is part of the solution, rather than part of the problem, in global efforts to achieve gender equality.
Building Partnerships to Advance Industry Change
As frontier technologies are transforming almost every aspect of our lives at un unprecedented scale, forging partnerships and building relationships with other stakeholders are essential to foster progress in a collaborative way, as well as to promote a more equitable and inclusive approach to these new technologies. With that in mind, UNESCO was planning a global conference on “Gender Equality and the Ethics of AI” that was due to take place on 7 March 2020, on the occasion of International Women’s Day. Due to the COVID-19 pandemic, UNESCO re-oriented this event to a global dialogue on gender equality and artificial intelligence with selected participants from the private sector, academia and civil society, focused on the design of gender equality principles for AI and their implementation.
In parallel, following the decision of UNESCO’s General Conference at its 40th session, UNESCO has embarked on a two-year process to elaborate the first global standard-setting instrument on the ethics of artificial intelligence. This inclusive and multidisciplinary process will include consultations with a wide range of stakeholders, including the scientific community, people of different cultural backgrounds and ethical perspectives, minority groups, civil society, government and the private sector.
In April 2020, an international group of 24 leading experts appointed by UNESCO held a virtual discussion to produce a first draft text for a Recommendation on the Ethics of Artificial Intelligence, with due consideration for various dimensions, including the environment and the needs of the global south.
This draft Recommendation is now published on the UNESCO website as part of a public online consultation. The purpose of the online consultation is to collect feedback from as many voices as possible to develop a comprehensive, inclusive and pluralistic text. The aim is to capture the interests, concerns and views of as many stakeholders involved in AI around the world, including the general public, academia, the scientific and technical community, civil society, the private sector, and governmental entities.