Mishcon de Reya page structure
Site header
Main menu
Main content section
The march of AI in the recruitment sector: are all uses good uses?

The march of AI in the recruitment sector: are all uses good uses?

Posted on 2 September 2019

Mishcon de Reya recently advised Impellam Group PLC on the acquisition of Flexy Corporation Limited.  Artificial intelligence (AI) has been strengthening its position in the recruitment sector over the last few years, and acquisitions such as this further demonstrate that AI is becoming a tool that recruiters should not ignore.

There is little doubt that AI, and in particular machine learning, can be a powerful tool when used in the recruitment industry. As Nick Davis, Head of Corporate and the Recruitment Services Group at Mishcon de Reya says, "The staffing companies that embrace technological change with an entrepreneurial spirit will be able to leverage their existing infrastructure, track-record and industry specific expertise to grow, shape and lead their industry."

However, as its use becomes more prevalent issues are coming to the fore.  Like all new technology, the use of AI in the recruitment sector needs to be carefully focused and harnessed to ensure that maximum efficiencies are achieved, and that implementation issues are avoided. 

One particular area of focus is CV screening. Manual CV screening is time-consuming, tedious and prone to human error. CV sifting tools have become highly efficient at scanning CVs to identify relevant experience, skills and even tone of voice.  Apart from driving efficiencies through automation, AI is designed to deliver data-driven objective decision making, freed from human bias and prejudice. However, unintended consequences have occurred in some implementation cases, for example when Amazon trained computer models to vet applicants using CVs submitted to the company over a ten year period.  Most CVs over this period came from men, reflecting historic male dominance in the tech sector, but in turn this taught Amazon's system that male candidates were preferable, an unintentional and undesirable consequence. All machine intelligence is built upon data that was initially created by people. Because of human bias and prejudice, the data and processes that inform AI can create inequalities in the system. 

Another example of unintended consequences occurred when a business started using an algorithm that was set to reject candidates whose English was poor.  Often these candidates were foreigners.  As a result, the algorithm, 'learned' that English sounding names generally meant acceptable qualifications and 'foreign' names did not – an unwitting, embarrassing and discriminatory outcome.

It is therefore the businesses that are working with the full knowledge of AI's capabilities and limitations that will do best.  As John Jersin, vice-president of LinkedIn Talent Solutions puts it, "I certainly would not trust any AI system today to make a hiring decision on its own, the technology is just not ready yet".

For now, AI can make hiring people faster and easier. Used in the right way it can help innovate, as Nick says, "We have seen a proliferation of exciting young companies seeking to transform the staffing industry with their innovative and technologically enabled solutions." However, it should not be seen as an end in itself.  The human element should not be lost in the mix, but instead AI should be harnessed to give recruiters back the time to enhance human connection and ultimately find the best people for the jobs available.  Algorithmic recruitment processes tend to focus heavily on experience and although that is not necessarily the wrong thing to do, it is also important to factor in a potential employee's personality.  Part of the challenge with using tech in the recruitment sector is that it reduces human contact.  People want speed and convenience, the demand for human touch is still just as strong.

How can we help you?
Help

How can we help you?

Subscribe: I'd like to keep in touch

If your enquiry is urgent please call +44 20 3321 7000

Crisis Hotline

I'm a client

I'm looking for advice

Something else