Josh Christianson is Co-Director of PEAT, overseeing its day-to-day activities and strategic initiatives. Prior to his role with PEAT, Josh worked at Deloitte Consulting, leading change management, technology and human capital initiatives at the Department of Health & Human Services as well as the Department of Veterans Affairs. While at Deloitte, Josh was a lead author of the report "Opening the Federal Talent Economy." Prior to Deloitte, Josh served as Career Program Manager for the Posse Foundation, a college access program focused on diversity and inclusion.
Discussions on Artificial Intelligence Dominate HR Tech 2019
The HR Technology Conference & Exposition (HR Tech) is one of the top global events showcasing emerging technologies and tools transforming the human relations (HR) industry. This year, platforms utilizing artificial intelligence (AI) continued to trend at HR Tech. The conference also infused a new focus on mitigating bias and ensuring that AI tools can align with corporate goals for diverse and inclusive hiring.
Our team observed three key themes for accessible technology and the impact of emerging technology on people with disabilities while participating in HR Tech 2019 in Las Vegas. These themes included AI activities at the forefront, C-suite support for access to emerging technology, and bridging gaps in inclusion in AI focuses.
Artificial Intelligence is Front and Center
HR Tech 2019 featured several sessions related to AI bias in emerging technology. Dmitri Krakovsky, the head of Hire for Google, emphasized that “AI will have the greatest impact when everyone can access it and when it is built with everyone’s benefit in mind.” Several other speakers at HR Tech 2019 outlined key strategies for mitigating the bias that AI can introduce into the hiring process. Himanshu Aggarwal, CEO & Co-founder of the AI-powered platform Aspiring Minds, described strategies for using AI responsibly.
Aggarwal and other speakers also noted that:
- AI measurements can offer broad indicators about a candidate’s qualifications, but humans need to be involved in the process for selecting data sets and evaluating the data to minimize and mitigate bias.
- The data measured and collected must be directly job-related; otherwise, companies may face increased risks for discrimination (intentionally or unintentionally).
John Sumser, Editor-in-Chief of HRExaminer (an online magazine), also illustrated the landscape of challenges presented by AI usage by adopting the metaphor of a fruit salad. He noted that AI programs can excel in identifying the individual ingredients that comprise the fruit said, while missing what he considers the best part entirely: “the mixed-up juice at the bottom.” Sumser also stressed that an AI system could mistakenly identify the polka dots on the fruit salad bowl as another ingredient. He offered this advice for working with AI-focused systems:
- AI is here, so start planning for it and include your legal team in the discussion.
- Be aware that machines may make major mistakes when using AI today.
- Machines can have biases pre-programmed in because they are designed intentionally with algorithms that can reflect the biases of their designers and developers.
C-Suite Support is Essential
HR professionals at HR Tech and elsewhere have increasingly viewed accessibility as a business imperative—but only when it is driven and championed at the executive level. Vendors at HR Tech noted that they only observed enthusiasm for accessibility from HR professionals themselves when corresponding commitment from leaders in the C-suite was also present. These experiences dovetailed with our team’s own understanding and insight into how to drive adoption of technology accessibility—that making forward progress requires solid support from top-level corporate leadership to be strongly effective.
Awareness of Disability Inclusion Lags in AI Discussions
When approached with questions about the AI behind their products, HR Tech vendors eagerly described their strategies for mitigating bias. However, most vendors only discussed how their strategies dealt with biases for gender and race. Few vendors showed sufficient knowledge of the issue to respond to this question in depth with respect to disability. We know from our work in promoting accessible workplace technology that people with disabilities represent a very heterogeneous group; this group contains many people considered outliers when compared with people without disabilities.
Thus, we find it very critical that these platforms pay careful attention to multi-factored dimensions of inclusion. We also find it imperative to recall wisdom from leaders like Jutta Treviranus, Director of the Inclusive Design Research Centre at OCAD University in Toronto, Canada. She notes that success in reducing disability-related biases in AI requires approaches rooted in the jagged starburst of human data—rather than simple bell curves.
HR Tech 2019 also reminded our team that awareness about the impact of emerging technology on people with disabilities remains the major immediate hurdle—especially when companies increasingly seek to benefit from inclusion of people with disabilities in the workplace. Unless HR systems align with diversity-focused hiring goals, companies will not fully realize the business advantages of hiring people with disabilities. In fact, new research from Accenture, Disability:IN, and the American Association of People with Disabilities showed that businesses leading in inclusion of people with disabilities saw a 28 percent gain in revenue. These organizations also witnessed a doubling of their net income and a 30 percent increase in economic profit margins.