“Facial expressions indicate certain emotions, behaviours and personality traits,” said Nathan Mondragon, Hirevue’s chief psychologist.
Umm... I think NOT.
Facial recognition technologies and Artificial Intelligence is not even close to being bias-free as researchers have shown time after time (see research from Joy Buolamwini for example).
These “AI-powered recruiting” technologies with bias embedded in them with *tested levels of inaccuracies* will not only replicate human bias at a much faster rate, but further exacerbate the systemic and institutionalized disparities experienced by already marginalized groups of people, namely people of color (especially darker skinned women, for whom facial recognition is the least accurate) and trans and non-binary folks.
This particular tech company claims it’s helping companies literally filter out people whose speech is slower, tone less desirable, and facial expressions not as positive.
The only success this company should be able to claim is that the technology has successfully learned and systematized racism.
AI is not bias free because humans that build it are not bias free.
Update: HireVue’s website says 700 companies use their product. The entire leadership team is white/white passing. Y’all.
#AI #diversity #facepalm