Women who don’t use a technology that increases productivity risk falling behind their male counterparts, but also risk being judged as having inadequate expertise. | Photo by Igor Omilaev on Unsplash.
What you probably already know: Women face harsher judgments than male colleagues for using AI at work. A study out of Hong Kong Polytechnic University and Peking University asked 1,026 engineers at a global technology company to evaluate a Python code snippet purportedly written by another engineer, either with or without AI assistance. While the code was identical across all conditions, the descriptions of how it was created differed. On average, reviewers gave 9% lower competence ratings to engineers they believed had used AI, even though the quality of the work was identical to the code created without AI. Female engineers received a competence penalty more than twice as severe as male engineers — 13% compared to 6%.
Why? Reviewers doubted a female engineer’s overall skills far more than a man’s when they evaluated the same AI-assisted code. This bias was particularly stark among male engineers who didn’t use AI themselves — they penalized female engineers who used AI 26% more harshly than men who did the same. And in cases involving AI-assisted code, reviewers estimated a larger contribution for the AI tool when a female engineer was involved. The study was initiated after company leaders tried for years to boost usage of a generative AI programming tool among software engineers through an incentive program. Despite their efforts, just 41% of the firm’s engineers used the tool — and the rate was far lower among women. “These findings reveal a paradox,” the study says. Technologies intended to enhance productivity may inadvertently undermine their users, impeding adoption and reinforcing inequality.”
What it means: While tech workers are more likely to use AI than people in other fields, it’s still not catching on as quickly as expected. A Pew Research Center survey found that just 16% of all U.S. workers used AI for work a little more than two years after the launch of ChatGPT. The results of the study at the tech company aren’t surprising — women are well aware of the competency penalty. Their reluctance to use AI tools, even when an employer encourages it for productivity, is a self-preservation instinct to protect their professional reputations in the face of existing gender biases. (They’re also concerned about the ethics of using AI tools.) “The AI assistance is framed as a ‘proof' of their inadequacy rather than evidence of their strategic tool use,” the study authors explain in the Harvard Business Review. In other words, those who might benefit the most from tools that boost productivity — particularly women in tech — are the ones who feel they can’t risk using them.
What happens now? Women face the competence penalty across the professional ecosystem, not just in tech fields. Past research suggests that women are viewed as less competent and committed than men, even when they overwork at the same rate, and figuring out a solution in which every employee can use AI safely is a delicate undertaking. This might mean reducing transparency when AI is involved. “While companies focus on access, training and technical infrastructure, they overlook the social dynamics that determine whether employees actually use these tools,” the studies authors say. Ultimately, employers will need to pinpoint penalty “hotspots,” such as the teams with the fewest women or a large number of men who don’t use AI. The most powerful champions are women in senior leadership roles who openly use AI, the authors say, because they lead by example and help influence skeptics to accept or — at the very least — respect it.
— Story by Cambrie Juarez
[email protected]
