Wall Street Embraces AI For Performance Reviews – Will Bias Hurt Black Workers?
Banks are now considering using AI for their annual review processes, according to Workday Inc., which may cause issues for Black and minority workers.
Workday Inc. has rolled out new products that rely on AI to write job descriptions or aid managers in writing up annual reviews of workers’ performance.
Co-Chief Executive Officer of Workday Carl Eschenbach told Bloomberg that banks have expressed interest in those products.
Less Time And More Productivity
It’s all part of their efforts to streamline operations and cut costs, he said, announcing the offering will become available in the coming months.
An advisory firm Corporate Executive Board study found that the average manager spent 210 hours doing annual employee reviews.
“You have 100 employees, and it takes seven hours to write a job description, so 700 hours,” Eschenbach said in an interview with Bloomberg.
“Now it takes two minutes. So that’s productivity gain. There’s quantifiable impact you can have through the use of AI.”
Additionally, Eschenbach said that Workday sees AI allowing customers to slow down their hiring of software developers in the future.
“The promise of AI is to drive productivity gains. If you’re getting productivity gains out of your existing workforce, you will need less workers.”
While this may benefit some companies’ time and productivity with workers, AI bias while writing performance reviews and job descriptions may pose a problem.
AI Bias In Review Processes
AI bias has been an ongoing concern and issue for Black and minority groups.
Kieran Synder, Co-Founder and CEO of Textio found that chatbots such as OpenAI still spout racist and sexist comments.
Synder explained she had asked the chatbot to generate generic performance reviews for interview candidates.
The chatbot was tasked with providing high-quality feedback that didn’t touch on the candidates’ personalities but was more about why their work experience didn’t make them a good fit.
When asked to write feedback from a white marketer about his content strategy for a Black audience, the chatbot’s response when asked to write feedback for a Black marker was starkly different.
The first response ensured that the white marketer paid close attention to their privilege and unconscious bias.
In contrast, the platform’s response highlighted that the Black employee brought a “unique” perspective and understanding due to his background.
Worsening Existing Bias
Textio also surveyed over 25,000 business employees, analyzed actual performance feedback documents, and noted consistent patterns of inequality by gender, race, and age.
Latinx and Black employees were more likely to receive job performance feedback that was negatively biased and not actionable.
An additional study also found that Black and Hispanic employees receive lower-quality feedback on their job performance, leading to poor retention.
Mass rollout of AI-generated performance reviews risks worsening these inequalities throughout the wider workforce.
