Facebook users may be seeing different job ads depending on their gender as the company’s ad-delivery algorithms can direct ads to a particular gender "beyond what can be legally justified" by differences in job qualifications, a new study has found.

Researchers from the University of Southern California studied Facebook and LinkedIn’s algorithms and found that in one of three cases that generated similar results, Facebook’s ad-delivery algorithms pushed an Instacart delivery job ad to a female-heavy audience. A Domino's Pizza delivery job ad was targeted to a male-heavy viewership.

According to the study, Instacart has mostly female drivers, while Domino's has mostly men.

"Facebook's ad delivery can result in skew of job ad delivery by gender beyond what can be legally justified by possible differences in qualifications," the study said, strengthening the argument that the social media giant’s algorithms may be in violation of US anti-discrimination laws.

However, Microsoft Corp's LinkedIn showed the Domino’s ads to nearly the same proportion of women as it did the Instacart ad.

"It's not that the user is saying, 'Oh, I'm interested in this.' Facebook has decided on behalf of the user whether they are likely to engage," said study author Aleksandra Korolova, an assistant professor of computer science at USC. 

"Just because historically a certain group wasn't interested in engaging in something, doesn't mean they shouldn't have an opportunity to pursue it, especially in the job category," the study added.

Facebook addressed the report in a statement released Friday.

"Our system takes into account many signals to try and serve people ads they will be most interested in, but we understand the concerns raised in the report," it said.

LinkedIn said the study's findings were in line with its internal review of job ad targeting, adding that the company recognizes that "systemic change takes time, and we are at the beginning of a very long journey."

US laws allow ads to be targeted based on qualifications, except in cases where protected categories like race, gender and age are involved. Despite tighter controls, researchers have expressed concern about bias in artificial intelligence software that chooses which users see an ad. 

see/sms (AP, Reuters)