I used to think AI bias was just a tech problem until a hiring tool in Atlanta rejected my resume

I applied for a project manager job last week and got an instant rejection email. The company uses an AI screening tool that flagged my resume for not having enough 'leadership keywords' from a specific list. It made me realize bias isn't just in the data, it's built into the rules we give these systems. Has anyone else had an AI tool misjudge their experience based on weird criteria?

4 comments

4 Comments

smith.parker3mo agoMost Upvoted

Yeah, that's a really good point about the rules being the problem. I had a similar thing happen where a system kept rejecting my friend's resume because she used "managed a team" instead of "led a team." The AI was literally just counting if the exact word "led" was there. It's not just biased data, it's also these overly simple, rigid checklists they program in.

sagecooper3mo ago

Wow, that example from @smith.parker really hits home. I used to think these systems just needed more data to be fair. But hearing it reject someone for saying "managed" instead of "led" shows the rules themselves are broken. They're not looking at what you actually did, just checking for specific words. That's not smart at all.

spencer4003mo ago

I read a case study from a university in Germany where they tested these hiring bots. The systems were missing qualified candidates because they only looked for keywords from past job ads, which were mostly written by men. So the bots basically copied old hiring mistakes instead of finding good people. It's like they automated the worst parts of HR. The fix isn't just better data, it's asking if we should use these dumb checklists at all.

willowg881mo ago

My grandmother got rejected by a similar system for a receptionist job because she wrote "answering phones" instead of "phone management." @spencer400 is right, these bots just copy the same old dumb rules and call it innovation.