AI recognizes your photo and reads emotions

2020-01-19

Artificial intelligence in recruitment keeps stirring up conversation. Element, an ATS (recruitment system), is building AI algorithms that automate parts of the hiring process.

Among other things, these algorithms can spot profile photos in application documents and read emotions from them. Element can tell whether a candidate is smiling, worried, or irritated, and gauge how strong those emotions are.

Face recognition in a CV document

Detecting photos in application documents has some obvious uses. When the system generates a candidate recommendation document, it automatically drops the recognized photo into the right place. Simple, saves time.

But reading emotions from a candidate’s photo? That one is trickier. Let’s set aside the extreme cases where someone looks unnatural or visibly upset in their picture. You could just ask them to resubmit with a better photo. I honestly can’t recall ever coming across one that bad in real applications.

So where does emotion recognition actually matter? Picture this: you’re hiring for a role that demands warmth or the ability to calm tense situations. Two candidates, same qualifications. One is smiling in their photo, the other isn’t. Does that tip the scale? Should it?

We keep bumping into questions like these as we wire AI deeper into recruitment. With Element, we want to build tools that help recruiters make better calls, not ones that make the calls on their behalf. ATS systems should back up recruiters, not sideline them.