As Meta prepares to make AI usage a core part of employee performance evaluations starting in 2026, concerns are rising about the policy’s unintended consequences for women in the workforce. While the company intends the new metric to reward innovation and efficiency, emerging research suggests that women may be disproportionately penalised for the very behaviours these systems aim to encourage.
The shift comes as major tech firms—including Microsoft and Google—also begin tying performance evaluations to AI adoption. But a 2025 study by researchers at Hong Kong Polytechnic University and Peking University points to a troubling pattern: when men and women use AI to produce identical work, women are consistently rated as less competent and viewed as making a smaller contribution. The bias is subtle but pervasive—women are perceived as relying on AI instead of demonstrating technical skill, while men using the same tools are seen as pragmatic and efficient.
This dynamic could worsen existing gender gaps in promotion, pay and visibility. Women already face higher scrutiny in technical roles and are underrepresented in engineering and product- leadership positions. An evaluation system that rewards AI impact but overlooks gendered perception gaps risks slowing their career progression even further. If AI usage becomes a “core expectation,” women could find themselves caught between adopting essential tools and being judged more harshly for doing so.
Meanwhile, organisations are rapidly expanding their AI monitoring practices. Worker-tracking software companies have seen increased demand from employers looking to assess AI adoption, meaning such evaluations may soon become standard across industries—not just in Big Tech.
Without safeguards, transparency measures, and bias mitigation training for managers, AI-based evaluations could disproportionately harm women’s workplace outcomes. As companies push for AI-driven productivity, acknowledging and addressing these gender-specific risks will be crucial to ensuring the future of work does not deepen existing inequalities.
