Research on open-source large language models (LLMs) revealed significant racial and gender biases in resume screening. The models strongly favoured white and male candidates, with Black men facing the most discrimination. These biases likely stem from societal privileges reflected in the LLMs' training data.

**For HR and recruitment professionals:** Your AI-powered resume screening tools may be perpetuating harmful biases. Audit your processes immediately. You're risking legal challenges and damage to your employer brand. Diverse teams drive innovation; by inadvertently filtering out minority candidates, you're potentially hampering your company's competitiveness. **For tech companies:** If you're developing or selling AI-based HR tools, you're now in the spotlight. Expect increased scrutiny from clients and regulators. You need to address these bias issues promptly or risk losing market share to more equitable solutions.

Expect a surge in demand for bias-auditing tools in the AI space. You'll likely see new regulations mandating fairness in AI-driven hiring processes. This could spark a wave of litigation against companies using biased AI systems. **For AI developers:** You'll need to invest heavily in developing more equitable training datasets and bias-mitigation techniques. **For all businesses:** Prepare for a shift towards 'explainable AI' in recruitment. You'll need to justify your AI-driven decisions transparently. This could lead to a hybrid approach, blending AI efficiency with human oversight to ensure fairness.