A new research paper explores the readiness of AI systems for use in hiring processes, specifically focusing on large language models (LLMs) for resume screening. The study developed a system architecture to separate candidate evidence, evaluation, permissions, and assessment records. Experiments with four LLMs across 210 resumes and six decision interfaces revealed that the effectiveness of AI in distinguishing candidates depends on the evaluation interface used, with scalar scores generally outperforming pass/fail judgments. The research also found that explicit permission rules and separation of resume content from instructions helped reduce unauthorized compliance, though LLM evaluators sometimes produced decisions inconsistent with their justifications. AI
IMPACT Highlights the need for careful system design and auditing to ensure fairness and security when deploying LLMs in sensitive applications like hiring.
RANK_REASON Academic paper detailing an empirical audit of AI systems for a specific application (hiring). [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX Code Finder for Papers
- CORE Recommender
- DagsHub
- Gotit.pub
- Hugging Face
- Influence Flower
- ScienceCast
- Spandan Roy
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →