AI-vision – A roadmap for diversity in tech talent recruitment

"... without acknowledgment from those at the top, change will be difficult to achieve."
9 November 2021

As one of the few black venture capitalists, British entrepreneur Eric Collins is determined to invest more in women and ethnic minorities — not just for reasons of principle, but for profit, too. (Photo by BEN STANSALL / AFP)

Over the past year, great emphasis has gone into sourcing and hiring new technology talent as resignations, retrenchments, and job migrations have become commonplace during the nebulous pandemic period. As organizations scramble to fill necessary tech job openings, the focus on diversity and inclusivity in teams has waned significantly.

As enterprises prioritize spending on tech enhancements instead of diversity hires, a recent report from Capgemini illustrated that while 85% of executives believe their organizations provide equitable opportunities for minorities, only one in five employees is female, and a mere one in six is from an ethnic minority background. 

Despite this, there is a strong movement advocating for more representation for other genders and races within the technology sphere. TechHQ spoke with Shobha Meera, the Chief CSR Officer for the Capgemini Group, to uncover the role technology is playing in creating both a more inclusive, and in some instances more divisive, workplace.

 

Why do you think such a gender-based gap is still so prevalent among IT teams today?

The gender gap in the technology sector is a complex issue. In part, it is tied to the fact that there is a limited pool of available talent. When we consider the number of tech and STEM students entering the field, the number of women in many parts of the world remains low, and the figure isn’t growing. For instance, only 21% of engineering major students are women, and only 19% of those studying computer and information science as a major are women.

The sector also has a perception problem, which is often exacerbated by tired diversity clichés that portray the IT and tech sector as male-centric environments where women and ethnic minorities are under-represented. While there is some truth here, the tech industry is vast, and progress in diversity is being made across different functions. For instance, women are progressing in AI, cloud, and software, but fields like engineering remain male-dominated. The perception issue extends inwards, as leadership figures and minority groups working in technology share wholly different views. 

The pandemic has compounded these issues. Women have been particularly challenged over the last 18 months; since the onset of the pandemic, 400,000 more women than men have left the workforce in the US alone. Estimates suggest that employment for women may not recover to pre-pandemic levels until 2024 — two full years after men.

There is an urgent need to address these inequalities. Without inclusive teams that emphasize diversity, organizations’ tech departments can’t build inclusive products and services, which in turn has an impact on customers’ experiences. Action starts with recognition. But without acknowledgment from those at the top, change will be difficult to achieve. 

 

What is the role of discriminatory technology in all this? Is it true that gender and ethnicity data can be used negatively against them, for instance not being given hiring opportunities?

Without inclusive and diverse IT teams, the risk of integrating human biases into new products and services is high. This has become particularly apparent when we consider AI-enabled systems. For instance, racial bias was recently uncovered in a major healthcare risk assessment algorithm used by over 200 million people in the US. The system reportedly classified Black patients as being at lower risk on average than white patients, despite the former having a higher incidence of chronic illness. Elsewhere, another AI hiring tool developed – and later abandoned – by a large global technology company was trained using a male-dominated dataset and excluded applications from two women’s colleges.

Our research revealed that many women and ethnic minorities have experienced tech-based discrimination first-hand, particularly when it comes to financial services products and healthcare. On average, 50% of ethnic minorities believe they were offered lower credit online for certain banking products, and 43% of women and ethnic minorities were either not shown, or denied access to, healthcare facilities in high-end locations or by those offering specialized services. What’s more, these groups are aware of the bias they experience due to technology and are becoming increasingly conscious of how their data is used, for fear of how it might impact them. 

 

What are the key factors to determine an inclusive tech team with emphasis on diversity? Does it extend to hiring policies?

To build a diverse team that represents society in all its richness, where everybody can thrive and be authentic at work, companies must cover the whole talent value chain. Organizations can start by creating their own sourcing pipeline through several channels: 

  • Act early: Engage young graduates or even undergraduates as a mid-term investment to benefit not only the organization itself, but the whole technology ecosystem. For example, run programs to attract young female talent, or educational projects for high school students.
  • Target untapped talent: Organizations can offer training for underrepresented and marginalized populations enabling them to acquire technology skills that are in demand in the job market. For instance, Capgemini’s Digital Academies initiative – aimed at those at risk of unemployment – trained 4,500 students in 2020 alone.
  • Adapt recruitment ads and organize dedicated job fairs: This allows every person to recognize themselves in expected profiles and skillsets.
  • Showcase inspirational role models: An effective way to do this is through internal testimonies. We recently launched a global podcast campaign, ‘HerStory’ featuring female employees who discuss their personal career stories, share tips and address how they were able to overcome challenges.
  • Ensure that your recruitment teams have limited bias: This can be achieved through Unconscious Bias workshops and training.

 

What are the building blocks to building an effective, inclusive team?

It starts at the top. When it comes to building an inclusive business, leaders must walk the talk, showing ambition and commitment. Leaders must also recognize the perception gap that might exist within their organization, between the leadership team and employees. This is the only way education and awareness building can start.

Clear KPIs should also be set and strictly monitored at both country and individual levels. This also means that HR processes should be streamlined and audited to ensure they offer equal opportunities in terms of staffing, development, promotion, and mobility. These efforts require a systematic approach: once an organization has attracted diverse talent they must nurture the pipeline from within, through educational opportunities, sponsorship of mentor programs, and networking.

Beyond process, inclusion must become part of the very culture of a company. Diversity Equity and Inclusion (DE&I) should be part of managerial training, and toolkits should be developed to enable managers to understand people with different backgrounds. Leadership teams should also drive collective mobilization: affirmative networks and employee resource groups for colleagues to engage with and contribute to make the diversity and inclusion agenda a reality.

 

What can companies that have an ethnic minority digital product discrepancy do to improve their output?

DE&I is not just an ethical consideration but a business imperative. Organizations can risk losing business if they do not address the demand for inclusive products and services from their customers. Today, not only are consumers aware of – and are experiencing – discriminatory technology, but they are also demanding action. 76% of ethnic minority consumers, and 61% of other consumers are calling for a change and expect organizations to develop technologies that can be used by a diverse set of consumers.

To limit bias and the risk of discrepancies, it is important to develop robust processes, practices and value systems that enable inclusion, and ensure women and minorities are fairly represented when building teams. Businesses need to work towards establishing the technological foundations for fostering inclusion.

They should consult with diverse user groups during pre-design and development phases to mitigate the risk of discriminatory technology. The current practice of purely asking inclusivity questions at the end of development cycles doesn’t cut it anymore. Businesses should also conduct impact-assessment analysis for algorithms and automated decisions, or screen datasets used to train AI systems for bias and audit them regularly as well as keep diverse users at the heart while developing inclusive digital products and services.