Including Women Key to Reducing AI Gender Bias: Google’s Bard, UNESCO & Experts

UNESCO will involve women in AI through the recommendation to advance the ethical development and deployment of AI for fair and inclusive results, with a specific emphasis on gender diversity and empowerment.

With the rapid growth of Artificial Intelligence (AI) raising questions on its impact on society, Google’s Bard, along with UNESCO and experts in India, believes that women’s involvement can help develop “fair, unbiased and beneficial to all AI”.

It was highlighted earlier that AI uses data produced by humans, but it also inherits human imperfections such as bias based on age, gender or ethnicity.

Google’s AI chatbot Bard shared an example of gender bias. It said if a dataset is made up of mostly men engineers, then an AI system trained on this dataset may be more likely to make decisions that favour men.

The chatbot then said: “Women’s involvement in the AI lifecycle can help reduce bias in AI systems in several ways. First, women can help identify and address biases in the data that is used to train AI systems. Second, women can help develop AI systems that are more inclusive and that take into account the needs of all users. Third, women can help ensure that AI systems are used fairly and ethically.”

In March, UNESCO’s Social and Human Sciences Sector launched the ‘Women 4 Ethical AI Platform’ during a high-level event. It was said: “This new platform will provide a unique space for global gender quality leaders in frontier technologies to combine their strengths and influence to achieve the clear goal of implementing UNESCO’s recommendation on the ethics of AI with a gender lens.”

UNESCO will involve women in AI through the recommendation to advance the ethical development and deployment of AI for fair and inclusive results, with a specific emphasis on gender diversity and empowerment.

The platform will focus on:

  • Dedicating funds to gender-related schemes
  • Ensuring that national digital policies include a gender action plan
  • Encouraging female entrepreneurship, participation and leadership in AI
  • Investing in programmes to increase girls’ and women’s participation in STEM and ICT disciplines
  • Eradicating gender stereotyping and ensuring that AI systems are not biased.

TIME FOR INDIA

Regarding the importance of involving more women in the field of AI, Neha Swetambari, SVP, Products and Delivery at Think360.ai, told News18: “AI allows for a micro-segmented approach to a huge base of consumers, like our country. Hence, India, while slow in adopting AI, will move at a tremendous pace. In this process, inherent biases will become a part of the system.”

She believes that having a diverse workforce will help identify biases faster and make them easier to rectify. “Biases introduced by low participation of women in the labour force or low ownership of property in risk frameworks need to be managed with equity-based weights to the right metrics and parameters.”

Sachin Arora, Partner and Head at Lighthouse (Data, AI and Analytics), KPMG in India, said AI literacy should be one of the core focus areas, as it will boost the GDP, given the productivity and jobs AI offers.

“We’re increasingly moving towards organisations becoming a collection of AI models. As we witness a significant shift, embracing diversity of perspective becomes even more paramount. AI teams building these models must represent the population they’re going to serve, otherwise, there is a huge risk of substantial bias affecting the outcomes of these models,” he added.

Meanwhile, Dr Mamta Arora, Associate Professor, Department of Computer Science and Technology, Manav Rachna University, also highlighted the importance of women’s inclusion, saying that gender biases in AI systems can perpetuate societal inequalities and reinforce stereotypes, leading to biased decision-making and unfair outcomes.

“India, like many other countries, is experiencing a growing interest in AI and its potential applications. We can promote women in AI by encouraging STEM education, women’s participation in AI research, supporting women-led AI startups, creating funding programmes, taking initiatives that inspire and support girls to engage in AI-related fields, and providing awareness of career prospects in AI,” Prof Arora noted.

Echoing a similar thought, Smita Khanna, COO at Newton Consulting Group, said that considering the rapid digitization, this is the best time to involve and encourage greater women’s participation in AI, which can be done via educational initiatives, scholarships, mentorship programs, outreach programmes and dedicated support networks.

She believes that AI can play a role in addressing gender-based violence in India. Citing the National Family Health Survey report of 2022, Khanna said that nearly one-third of women have experienced physical or sexual violence.

“Indian non-profit organization Safecity uses AI to collect and analyze data on incidents of sexual harassment and violence against women. This data is then used to identify patterns, inform policy advocacy and implement preventive measures in communities,” she noted.

Leave a Reply

Your email address will not be published. Required fields are marked *