In part one of this article, we noted that artificial intelligence (AI) is being used more and more throughout the recruiting process. We talked about some of the many examples of that and started to review some of the benefits. Today, let’s further the discussion by outlining some of the potential drawbacks.
Source: Paul Campbell / iStock / Getty Images
Using AI in the Recruiting Process: Cons
Here are a few of the potential drawbacks to utilizing AI in the recruiting process:
- Applicants may not love interacting more with things like chatbots (recruitment bots) and the like. The lack of personal interaction may turn some people away.
- Putting hard rules in place to screen applications may create too much rigidity, thus possibly overlooking candidates who have nontraditional backgrounds but would otherwise be perfectly qualified and able to do the job. It can also overlook otherwise-qualified candidates who simply did not use the keywords that are being targeted. In short, great candidates could be missed because of the rules AI follows.
- Conversely, the system can easily get things wrong if it screens for keywords without context, resulting in finding applicants who actually are not qualified. Or, it may select applicants who used the right keywords but are not a good match or are trying to deceive the system.
- AI doesn’t always eliminate biases because it still has to be taught what to look for—so biases can still come into the process inadvertently. (See, “The Amazon Example: Can AI Discriminate?” for a big example of AI-embedded discrimination in recruiting.)
- AI can seem impersonal, which can negatively affect company perception and the employment brand.
- Applicants may fear their personal data could be compromised, especially when the process seems less personal to begin with.
- AI can take a lot of time to set up properly, especially when trying to get it right and minimize the downsides.
- A chatbot cannot answer complex questions, and users may become frustrated with it.
- Even good AI-based screening systems have difficulty evaluating soft skills, which are crucial to many jobs. (It’s not impossible, but it’s harder to assess.) It’s not simple for AI to assess cultural fit for the organization either.
What has your experience been with adding more and more AI into the recruiting process? What have you learned along the way?