The Future of AI in Recruitment
Technology is fast changing the way we work. In many instances, solutions like artificial intelligence (AI) promise to optimize processes, automating monotonous tasks to make things more efficient and allow employees the time to do more value-adding work. Good for some— not so much for others. In fact, a study by Brookings estimates that some industries, such as transportation, may lose up to 70% of jobs to automation.
The recruitment industry has seen its share of automation in the last few years, and we have seen machine learning solutions introduced in some areas with (in my opinion at least) decidedly mixed results. Whatever the outcomes, however, AI is here to stay and is moving at an incredibly fast rate, so it’s worth considering the benefits and challenges.
Large corporations like Walmart, FedEx, and American Airlines currently use machine learning tools to streamline their hiring processes. AI platforms are said to offer a “dependable” and “bias-free” way of matching employees with employers. These companies use a platform that analyzes answers to interview questions, using words and language to put together a profile of an individual and assess their suitability.
Recruiters have been using AI for longer than you might realize. For example, key word searching. Technology is used to screen resumes for things like keywords and skills listed on the job spec, qualifications, etc. It’s potentially useful, particularly when you expect a job to attract a large number of potential candidates, but those who miss using those words, however suitable they may be for the role, often end up falling at the first hurdle.
AI, therefore, can hold a useful and important place when screening for more entry-level positions. A competitive market full of recent grads who tend to be similarly qualified and all vying for the same position means some automated way of filtering through that first stage is an advantage for recruiters and hiring companies. AI programs are now even being used to conduct virtual interviews at this level and can make initial decisions faster than an HR manager.
After this initial process, however, it’s generally time to entrust a recruiter. Humans can identify and extract information in an interview that a machine simply cannot. Being able to contextualize a candidate’s skills and experience and paint a picture of how the individual would do the job—not how they have done their previous job, but based on fundamental competencies, soft skills, and life experiences they have gained—can simply not be measured by a machine just yet.
There is also the issue of diversity and inclusion—blind hiring and using technology in the early stages can go some way in eliminating unconscious bias. Still, AI is not the holy grail of impartiality many hoped it would be. In 2015, Amazon found itself in hot water as its AI solution was automatically rejecting resumes from women for technical roles. The machine was making decisions based on hires the company had made over the past ten years, and these ended up being predominately male. While many people perceive AI as “thinking for itself,” it’s actually just a machine making decisions based on data—if the people inputting that data have unconscious bias, so too will the machine.
These issues and more show AI is unlikely to replace the recruitment process any time soon, but it will certainly change the way it looks. It’s essential that recruiters, companies, and candidates learn how AI can be introduced to processes in a way that complements, improves, and augments, but doesn’t replace the human element of the process. Those of us in the industry will need to pivot slightly as parts of the job we do are replaced by bots, but other aspects will become even more critical. As the role of a recruiter transforms, our success will rely on an understanding of what a computer can and can’t do, as well as the support that companies and candidates need from us.