3 Key Risks Your AI-based Recruitment Strategy Has Hidden Within
Technology has been adopted in practically every area of life, from home management to the way we work and do business. In recent years, HR has made use of Artificial Intelligence (AI) in their recruitment strategies. But while that move has brought clear advantages, there are some risks to be aware of.
Of course, technology is almost always an aid that greatly improves the cost and time efficiency of a process. When it comes to identifying new talent in the jobs market, AI has proved to be a major help, particularly when applied to a suite of human resources management systems (HRMS). Machine learning techniques allow this software to screen CVs and assess suitability in seconds.
However, it is not unheard of for technology to fall short, and there are some areas where software is unable to make allowances where the experience and foresight of a HR professional can.
In October last year, it was reported that Amazon had scrapped its AI recruiting software because of its apparent bias against women. The algorithm was based on resumes from the previous 10 years, which meant there was a male-centric standard that female applicants were being assessed against.
3 Risks To Using AI In Your Recruitment Strategy
-
- Hiring Via Trend-Based Projections
Many of us live by the old adage: ‘if it’s not broken, don’t fix it’. When it comes to recruitment, replacing departing staff is made easier by bringing in a like-for-like replacement. AI-based software looks at the historical trends for the job in question by examining the CVs of previous personnel and then searching for the most similar CVs from prospective candidates.
But such a strategy can only bring success if factors in the sector remain unchanged, and this is a highly unlikely scenario. Constant changes in industries, markets and technologies mean job descriptions are constantly evolving. Analysing historical trends is not enough; recruiters have to have one eye on future trends too.
-
- Human Experience
AI relies on the quality of data and the algorithm it’s given to mine through it. While this is fine for identifying the relevance of past positions, it doesn’t cover such critical aspects as the Human Experience. By this, we mean experiences that play a role in forging and showing such important personal traits as attitudes, values and behaviour.
Everyone is unique, so the context a person works in can be telling. A new candidate may have the same education and worked in similar roles as an existing top performer, but data does not reveal his or her ability to plan, perform and deliver. AI-based recruitment software can easily overlook a candidate whose attitudes are ideal.
According to a recent Yoh survey, 69% of US workers think AI “has no place” in hiring tasks, and 33% believe AI is “not appropriate in assessing candidates’ truthfulness about their experience and qualifications”.
-
- Keyword-Heavy Searching
Lastly, can AI really match the perfect CV with the Job Description? The algorithms used are word dependent, identifying the best candidates based on the words they used in their CV. Some experts argue that this creates a problem where savvy candidates will come up tops because their application included the right words enough times.
In fact, the most important properties a candidate should have are usually mentioned only a couple of times, making it very likely they will be included low on the shortlist, if included at all.
This is actually counter-productive, undermining the whole idea of adopting AI-based recruitment software in the first place. To be thorough, several screenings will be necessary, which only increases time and cost of the exercise.
