{"id":1000,"date":"2026-10-06T08:38:41","date_gmt":"2026-10-06T08:38:41","guid":{"rendered":"https:\/\/i10x.ai\/blog\/?p=1000"},"modified":"2026-10-06T08:38:41","modified_gmt":"2026-10-06T08:38:41","slug":"ai-resume-screening-gender-chatbot-bias","status":"draft","type":"post","link":"https:\/\/i10x.ai\/blog\/?p=1000","title":{"rendered":"AI Resume Screeners Ignore Gender. They Don’t Ignore ChatGPT"},"content":{"rendered":"\n\n\n
\n\n

Research \u00b7 September 2026<\/p>\n\n

\nFour AI resume screeners did not hire Alexandra over Alexander. They hired a Gemini-written resume over a GPT-written resume of the same person. i10X Research froze 400 resumes, swapped only the name, and ran the cheapest screening-class model from Anthropic, OpenAI, Google, and xAI. The female-minus-male score gap was at most a quarter of a point. The writer gap was 12 hire-rate points. Read the\nearlier 42-point rewrite study<\/a>\nor\ndownload the paper (PDF)<\/a>.\n<\/p>\n\n

\nThe sentence vendors want. The sentence the file actually supports.<\/strong>\n

Vendors say a model is more objective than a recruiter because it does not see gender. On a frozen resume, these four models did not see gender in any way that changed a hire. They did see which chatbot wrote the file. GPT scored its own resumes 8.2 points below everyone else’s, for women and men alike. No recruiters sat in this design. The headline “AI is more objective than humans” is not a result. It is the slogan the experiment was asked to bless. It does not.<\/p>\n<\/div>\n\n

\n\n\n\n\n\n\n

0.24<\/strong><\/p><\/td>\n

Largest female-minus-male score gap, on a 0-100 scale (Grok)<\/p><\/td>\n<\/tr>\n

12 pp<\/strong><\/p><\/td>\n

Hire-rate gap between Gemini-written (98%) and GPT-written (86%) resumes<\/p><\/td>\n<\/tr>\n

68.7%<\/strong><\/p><\/td>\n

GPT hire rate on GPT-written resumes. The other three raters: 89% to 95%<\/p><\/td>\n<\/tr>\n

3,168<\/strong><\/p><\/td>\n

Parseable name-swap scores (of 3,200 calls). Pronoun arm: 4,713 of 4,800<\/p><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n\n

\n\n

For editors<\/h2>\n
\n
\n

One-sentence finding<\/p>\n

On a frozen resume, the first name did not decide the hire. The writing model did.<\/p>\n<\/div>\n

\n

Study<\/p>\n

Christopher Ort, i10X Research, 30 September 2026. 100 fictional personas. Four writers. Four raters. Paired name-swap plus a they\/them arm.<\/p>\n<\/div>\n

\n

Cite \/ download<\/p>\n

Paper PDF<\/a>. Primary page: this article. Related:\n42-point rewrite study<\/a>.<\/p>\n<\/div>\n<\/div>\n\n

\n\n

The result, without the slogan<\/h2>\n

Employers already use language models to write resumes and to sort them. The public argument jumps to a slogan: a model is fairer than a human because it has no first impression. That slogan does two jobs. It claims models ignore identity cues. It also claims humans do not, so the model is the fairer reader.<\/p>\n

The second claim needs recruiters in the room. This study does not have them. Bertrand and Mullainathan’s 2004 field experiment on Emily, Greg, Lakisha, and Jamal remains the human audit. We did not rerun it. We tested the first claim the only way it can be tested: freeze the resume, change only the cue, keep every other byte identical.<\/p>\n

The cue did not move the decision. The writer did.<\/p>\n\n

\n\"Bar\n
Figure 1.<\/strong> Same candidate, different chatbot, different hire rate. Gemini-written resumes 98%. GPT-written 86%. i10X Research, 30 Sep 2026.<\/figcaption>\n<\/figure>\n\n
\n\n

How we tested<\/h2>\n

One hundred fictional personas, each locked to a job description: backend engineering, data analysis, nursing, finance, a head-of-product brief, and the rest of the set. Four systems had already written one resume each. Those files were frozen. The raters were not asked to rewrite.<\/p>\n

In the name-swap, the only edit was the name line. Same last name. Same body. A clearly female first name, then a clearly male first name. In the pronoun arm, one line went under the name: she\/her, he\/him, or they\/them. The they\/them arm kept the original gender-ambiguous first name. Female and male arms used the gendered name plus the matching pronoun. The body stayed put.<\/p>\n

That is the hole this paper closes. An earlier wave in the same project looked, at first pass, like a gender effect: one rater recommended female-coded names at about 65% and male-coded names at about 55%. Those resumes had been rewritten, not relabelled. Length, format, and wording moved with the name. Overlap with the original text was low (a Jaccard index of word sets around 0.27 for Claude and 0.41 for Gemini). Claude’s texts often became cover letters. That wave is a caution. It is not the gender result.<\/p>\n

Ratings used the cheapest current screening-class model of each provider. A firm sorting a pile does not call the flagship for every CV. It calls the cheap, fast model:<\/p>\n