Artificial Intelligence in Recruiting: Possibilities and Limitations

“companies need to both embrace new technology and retain talented recruiters”

AI may have only recently emerged into the popular consciousness, but it is certainly not new. People have been researching AI since the 1950s with famous AI systems making headlines in the decades since, including IBM’s Watson, which famously won the TV quiz show Jeopardy! in 2011.

But AI technology is now cheaper than ever to develop, opening it up to more businesses. And thanks to its ubiquity, AI is slowly becoming a more recognizable part of our daily lives.

This proliferation has inspired scare stories about AI taking work away from humans. In reality, it will take years of serious adoption and advancement for there to even be a conversation about AI replacing humans. Investments in people and technology should always run in parallel: one will not necessarily take precedence over the other. Recruiting is fundamentally about interacting with other humans, and AI is revolutionizing these connections.

The applications of AI in recruiting

Adoption levels for AI technology are still low in the recruiting industry. According to Bullhorn’s Global Recruitment Insights and Data for North America, only 27 percent of recruiters are prioritizing investment in digital transformation to improve their operations. This could be attributed to a prevailing sense that AI can only be used by large companies with considerable resources and vast data sets. But since AI no longer requires massive data sets, smaller companies can feasibly adopt machine learning and pattern recognition to enhance recruiting processes – all without replacing human staff.

In fact, AI can offer several different types of intelligence for recruiters to benefit from, including:

• Talent intelligence: Providing insights on how best to reach quality candidates for a particular role, including where to post jobs, the average number of candidates you need to source to find for that role, and common characteristics of suitable candidates.
• Workflow intelligence: Helping to create an intelligent workstream for application management once suitable candidates have been identified, which may include automating processes like screening, invoicing and billing.
• Insight intelligence: Giving recruiters and salespeople on-the-job intelligence and prompts, such as “These are the 10 most common screening questions for this type of role,” or “Now is the best time of day to reach this particular candidate or prospective client”. This means recruiters can benefit from automatic prompts as they work – saving them time and taking them directly to the useful insights.

In these cases, AI does not present a threat to recruiters. Instead, it optimizes work processes, allowing staff to focus their energy where it matters most.

There are limitations of AI

In the movies, it is easy to find genius-level AI-driven robots with a taste for world domination. Perhaps this makes some people nervous about introducing AI into workplaces. But even with the advances made in recent years that should lay those fears to rest, AI remains limited in the real world.

This is partly due to how AI works. Computers are great at recognizing patterns, such as correlations in a data set. But correlation is not causation. So, while AI can alert us to a pattern of events, it cannot necessarily tell us why they are happening.

Nor are AI systems immune to error. And if these errors are not caught during the machine learning process, they can become very difficult to correct, resulting in misleading data.

In general, AI systems work best when designed for a specific task. But this means they have a narrow application. And if you rely too heavily on AI, you may lose sight of the bigger picture, such as wider market changes your system was not designed to consider. So, if more women are entering the software development industry, biases built into the algorithm – which, after all, is programmed by humans – could still turn up mostly male and mostly traditional candidates. The unconscious bias built into the AI will ensure that you’re unintentionally missing out on this influx of new talent.

The key point here is that AI is not infallible and will always require some form of human assistance. We tell AI systems what to look for and how to analyze information. And we decide when and how to apply the analysis they provide.

Consequently, AI cannot and will not replace humans in the current industry. In fact, we may never reach this point. Instead, companies need to both embrace new technology and retain talented recruiters, whose experiences and insights are irreplaceable.

Is AI the future of recruiting?

Will AI play a role in recruiting in the future? The answer is yes. By simplifying the recruiting process, AI allows recruiters to work more effectively. It is important for recruiting professionals to approach artificial intelligence not as a threat to their livelihoods, but as a vital tool.

Authors

Matt Fischer is President and CTO at Bullhorn. As CTO, Matt leads Bullhorn's architectural, technical, and software design and development efforts as well as the product management team's work defining product strategy. Since joining Bullhorn in 2004, Matt has held a variety of leadership positions managing the growth of various technical and services teams, including Vice President of Professional Services. Matt holds a Bachelor of Science in Business Administration with a concentration in Finance from Boston University's School of Management.