Call us now: AUS 1300 878 473 | NZ 0800 000 849
Signup   | 

AI in Recruitment: Where are we?

Too many acronyms… HELP!

AI:

artificial intelligence, AI is an area of computer science focused on creating intelligent machines (‘robots’) that can think and behave just like humans.

For instance: machines may mimic human processes like planning, solving problems, understanding speech, and even operating cars! ‘Mainstream’ examples of AI include drones, driverless cars, and ‘home assistants’ like Google Home or Amazon Alexa.

AI.jpg

ML: Meaning machine learning, ML is a specific application of AI (above). Using statistical methods, it trains intelligent machines to automatically learn from data. They can identify trends and patterns in the data, and then make predictions based on these.

A basic example of ML is Netflix’s ‘recommendations’ section: it ‘learns’ from content you have previously watched, and uses this data to predict similar TV shows or movies you might enjoy.

A.I in recruitment

Making recruitment decisions is a highly time-consuming process. It’s no surprise, then, that researchers have looked to AI to help make some of these decisions for us! AI can automate and streamline some parts of the recruitment process, potentially saving considerable time and money.

 

Current applications of AI in recruitment include:

·       Using language algorithm software to identify biased language in job advertisements

·       Using ML to screen resumes & compile candidate shortlists

·       Using ‘data scraping’ programs to analyse a candidate’s LinkedIn profile

·       Using recruiter ‘chatbots’ to answer candidate queries in real time

·       Using ML to rank and shortlist candidates, based on data from existing high-performing employees

·       Assessing candidates’ choice of words or speech patterns in video interviews

·       Using AI software to schedule interviews


Some recruiters report that AI tools help to free up their time, so they can focus on the more ‘human’ elements of their roles. We believe that a primary benefit of AI is its ability to automate repetitive, high volume administrative tasks. Innovative tools that simplify and speed up the recruitment process are likely to be attractive to many organisations. It is also true that in some scenarios, algorithms have been demonstrated to outperform human experts.

Proceed with caution

For all of its benefits, however, AI is not without its flaws. Being a diversity-centric organisation, Testgrid’s main concern is that AI can learn human biases.

Because intelligent machines ‘learn’ from data created by humans, they are vulnerable to acquiring unconscious biases already present in society. Language inherently contains biases, and so machines trained using language cannot avoid picking up these biases. In a nutshell, any bias that already exists in your recruitment process can, unfortunately, be learned by AI.

An example of this might be using AI to scan initial job applications and compose a candidate shortlist. But what if your AI tool has ‘learned’ from the resumes of existing staff, and these employees all happen to be white men who attended a certain university? Whilst not intentional, the algorithm has learned that these characteristics are associated with success. There’s even a term for biases being hardwired into an algorithm: ‘algorithmic bias’. This will not fix our diversity problem. In addition, AI development is currently a male-dominated industry, and consequently the technology can reflect this bias.  

Amazon learned this the hard way recently: they found that their ML tool (which had been 4 years in the making) discriminated against women. Their tool had analysed applicant data submitted over a 10-year period, deduced that most of these data scientists and engineers were male, and learned that male candidates were preferable.

Consequently, we recommend caution when selecting AI-driven recruitment tools. While AI has its merits in certain recruitment contexts, we are still only just beginning to learn about its potential flaws in recruitment settings – bias being one of these. As with all technology trends, we are interested to see how AI in recruitment develops over the next 12 months. At present, it appears that the value of AI lies in supplementing, not replacing, human capability.

Interested to learn more about Testgrid’s best practice recruitment tips? Speak to one of our experts today!