KI-Einstellungsalgorithmen lehnen schwarze, asiatische Arbeitssuchende häufiger ab
https://www.theregister.com/ai-ml/2026/05/27/ai-hiring-algorithms-reject-black-asian-job-seekers-at-higher-rates/5247387
KI-Einstellungsalgorithmen lehnen schwarze, asiatische Arbeitssuchende häufiger ab
https://www.theregister.com/ai-ml/2026/05/27/ai-hiring-algorithms-reject-black-asian-job-seekers-at-higher-rates/5247387
20 Kommentare
Because the original human-based hiring system was already biased so the algorithmic system does the same?
*Na-uh, it’s not discrimination. It’s AI!*
I would like to see more data on what AI also looked at.
For example – level of education completed/ high school diploma or GED and so on.
I bet there is a correlation there as well.
They also tend to reject anyone over 40. Hang on to your jobs kids!
Wonder how the AI being used determined the ethnicity of the applicant. I would think that information and the actual name and birthdate would not be entered.
> AI hiring algorithms reject Black
**PHEW** thank god it’s not woke! You should have to earn your..
>, Asian job seekers at higher rates
Wait. wtf?! That’s not what I meant when I said I hope AI is anti-woke!!
Back in college, I received a surprise 400-dollar settlement check due the company discriminating against black and Asian candidates. I received an internship offer but still got the check lol.
Shocked Pikachu face.
it s just copying its training data
There shouldn’t be hiring algorithms…
You mean like humans do? Yes, that’s the model. Don’t expect more from AI than humans,
hope that asian guy that did all that for affirmative action realizes he was pawn
Le racisme 3.0 est bien réel c’est une folie anti sociale
This is just another artifact of non technical leadership rushing to capitalize on AI without fully understanding its risks. „Give me only applicants who do not need an H1B“ might filter out people who have „H1B required“ in their resume, or it might start throwing people with foreign sounding names in the garbage depending how your platform is designed, governed, and prompted.
„The game-playing approach used by pymetrics may lack that sort of demographic information, but the researchers say they find adverse impact despite the absence of demographic details and pymetrics‘ efforts to de-bias applications.“
It seems that the model lacks data for discrimination and the article’s conclusions are far-fetched.
Ai doesn’t hire job seekers. Hiring managers do.
Also, correlation doesn’t equal causation. It may be that AI algorithms are rejecting resumes from Black and Asian job seekers at higher rates, but that alone isn’t proof that their resumes are being rejected just because they’re Black or Asian.
Instead of crying foul, maybe it’s for a reason.
More loyal? Better performance? Less trouble? I mean, humans and the AI training data go hand to hand. Stereotypes are very real.
I may sound countroversial, but…
> They made that determination by applying the US Equal Employment Opportunity Commission’s „four-fifths rule,“ which at least on paper elicits agency attention when a given group’s hiring selection rate is less than 80 percent of the most recommended group of job applicants.
It seems that the platform did not account for applicants‘ races and for that has been deemed racist.
I don’t know what to say, honestly I don’t. I thought that it was racist to use race as a factor during hiring.
Oh, to have a functional Department of Justice and Consumer Protection Bureau. 😑
Things like this remind me how truly fucked up our government is right now.
Blind hiring also „discriminated“ heavily, or at least resulted in even sharper miscorrelations. Perhaps the reason isn’t just bias or discrimination, and has something to do with objective credential and work experience differences?