It is interesting to see how as each new year rolls by a new “hype” comes with it – and “AI” is the current subject that dominates the conversation both on and offline. But unlike its predecessors, AI seems to have gone from hero to zero in in less time that it takes to say “the next big thing”.
A significant amount of the noise surrounding AI in the sphere of HR is around bias in the hiring process. More specifically, the application of “AI” based screening or sourcing tools in the hiring process and their potentially discriminatory impact. This debate gained momentum a couple of years ago in the wake of Amazon’s “bias recruiting algorithms”.
HR practitioners, recruiters, influencers and guru’s alike have been quick to put “AI” (or perhaps more accurately, algorithm based screening technology) in the dock on serious charges, but I wonder if this is simply a bunch of folks in glass houses lobbing stones. Notwithstanding that the technology available now is entering a level of sophistication that we should indeed be handling carefully, I do feel that we are somewhat shooting the messenger here.
The strength of feeling and commentary around this issue would lead you to believe that before technology, everything was unicorns and princesses in the world of selection and hiring. That, until we unleashed the first algorithm, the playing field of recruiting was perfectly level. That our selection and interviewing practices were bias free. This is simply just nonsense.
The inter web is chock full of reports discussing the lack of diversity in organisations, under representation of minority groups in senior level or technology roles, ethnic minorities being excluded from shortlists or hires and judgements being made around sexual orientation of applicants. This is not new news.
And these reports are often shared by the very same people who are quick to point out the shortcomings of the technology whilst refusing to acknowledge the root cause of lack of diversity in organisations today – you and me. And every hiring manager, recruiter, HR professional that ever walked the earth.
The simple truth is that as every day goes by we continue to make poor judgements, unleashing our unconscious and conscious biases yet remaining completely silent on our own complicity, instead choosing to deflect responsibility and point the finger at the very technology that is highlighting the issue.
I am also concerned about our obsession with technology and how we think it has all the answers. One of the most common requests for technology in our sector is to “blind” all applicants – remove any information from the application that would give a clue to their gender, ethnic origin or age etc. It’s seen as an important feature and standard fare in the diversity recruitment technology landscape.
Whilst we might see this as good intentions, isn’t it simply kicking the can of bias down the road? The second question I get asked the most would attest to this: “Can you help with removing bias further down the funnel, at the interview stage?” This question just demonstrates that as much effort we might put into trying to level the playing field through anonymising data at the top of the funnel, the moment the interaction gets human, the real bias creeps in.
Since when was it deemed acceptable to abdicate our responsibility as recruiters to technology? Why is outsourcing our judgement on what is such a critical issue accepted without question? And why do we find it acceptable to completely ignore the real reasons behind the lack of diversity whilst shooting the messenger that comes bearing the news we all can’t seem to face.
Improving our appalling lack of inclusivity and representation in hiring and in the workplace in general requires a change in behaviour. And a change in behaviour requires a shift from ignorance/denial into acceptance. There is most definitely a place for data science to play in surfacing issues and giving us very useful insight into how we can shift the dial.
Ultimately though, breaking the cycle of exclusion in the recruitment process and the workplace is a human issue, not a technological one. Putting the technology in the dock and abdicating our personal responsibilities to address bias to a technology solution isn’t the answer.
Endless debate about the “fairness of AI” and bias will change nothing unless we acknowledge the root cause of the issue – Us.