With the rise of Artificial Intelligence in everyday tasks, it comes to no surprise that AI is not only being used by companies in their hiring process, but they’re doing so at an alarmingly high rate.
It also comes to no surprise that the companies who use AI to scan their resume applications argue that its use is a good thing because it reduces human bias.
And a lot of how AI works in these processes is based on the human inputs themselves, which presents a huge issue with AI, as was found in one report where it found “most hiring algorithms drift toward bias by default.”
It was also found this how AI learns the traits of a ‘good hire’, which are typically based on past racist and sexist hiring decisions.
We know bias inputs lead to bias outputs, so wouldn’t the obvious solution be to program AI to ignore those past hiring decisions, and to, you know, not be so bias?
This can prove to be more difficult than stated, with AI already finding workarounds to get those same results, as was found in another report, this time with Amazon’s hiring AI.
It found that Amazon’s hiring tool taught itself to prioritize male candidate resumes,
while penalizing resumes that included the word ‘women’s’ and downgraded resumes of graduates from two all-women’s colleges.
And in yet another company, its hiring algorithm identified two factors to be the most indicative of job performance
when scanning the resume applications they received. Those two factors were if your name was Jared and if you played high school lacrosse.
So, suffice it to say:
The data inputted into hiring AI, and the outcomes they’re trained to prioritize, matter significantly.
With the increased use of AI within the hiring process, and with little to no government regulation in place to ensure these programs are giving every candidate a fair shot of having their resume seen by a human, these programs have the ability to create further discrimination within the hiring process.
Comments