Artificial intelligence job screeners prefer AI-written resumes over human ones, researchers find

When choosing a bot, choose wisely.

Job seekers now face a modern challenge: applicant tracking systems that prioritize candidates who use AI to write their resumes.

A study revealed that AI-powered applicant tracking systems prefer resumes written by AI over those written by humans, especially if they used the same large language model as the company.

According to the study, “AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights,” LLMs tend to favor resumes they generated themselves over human-written equivalents.

This bias, identified by researchers from the University of Maryland, National University of Singapore, and Ohio State University, could disadvantage strong candidates.

If this bias is not addressed, it could impact hiring outcomes by favoring candidates who use the same LLM as employers, while overlooking equally qualified applicants who do not.

Emma Wiles, a Boston University Professor specializing in information systems and AI’s impact on the labor market, highlighted that AI tools should complement human writing, not replace it.

The researchers analyzed 2,245 human-written resumes and found that AI evaluators were more likely to select candidates who used the same LLM, such as GPT-4o and Deepseek-V3.1.

The study revealed that the bias was most prominent in accounting, sales, and finance.

This preference for self-generated content raises fairness concerns for job seekers and poses risks for employers by potentially overlooking qualified candidates, introducing a new form of bias in the hiring process.

More than 300,000 job cuts were announced between January and April 2026, particularly in the tech sector, according to a report by research firm Challenger, Gray & Christmas.

Wiles advised workers to use tools that enhance human writing rather than overshadow it, ensuring that their true selves shine through in their resumes.