Artificial intelligence is opening doors to many opportunities for entrepreneurs. But like many technologies, not every one of those doors has a great prize behind it.
It seems AI is everywhere, but one notable place where AI has a growing influence is in the recruiting and hiring process, and this can be a two-edged sword.
There is an inherent tension to the hiring process: Job seekers want jobs, employers want the best employees, and it’s a big decision for all involved. Thus, it is entirely natural for prospects to do everything possible to make themselves look good, and for companies to probe each candidate thoroughly.
“Creating a good experience for candidates while meeting your business objectives is a delicate balance,” observed Chris Frost, director of client optimization at HireVue, a human resources company. “Candidates need to feel like the barrier isn’t too high to participate in the selection process, and employers need to gather enough information to make an informed decision.”
Jobseekers are increasingly using AI to assemble their resumes and prepare for job interviews. Within certain boundaries, this is completely logical; given that AI is a tool being used ever more widely, a prospect’s ability to utilize it effectively could well be considered a plus.
Those boundaries are crossed, however, when candidates misrepresent their accomplishments and expertise, when AI is used to generate answers to an employment questionnaire (as opposed to simply using AI to polish the answers), or when those answers are outright copied from another source. In its most egregious form, a dishonest candidate might even have AI complete an entire pre-employment assessment.
According to Frost, this type of cheating remains rare, though he noted that HireVue is “engaged in ongoing monitoring through automated techniques in interviews, as well as more detailed/manual review of VJT [virtual job training] responses.”
For entrepreneurs concerned about potential misuse of AI, Frost suggested several ways to deter and/or detect it. This starts with how the hiring and interview process is designed, particularly the questions asked of candidates.
“Questions requiring situational judgment, personal experience or explanation of decision-making are far harder to fake,” he noted.
Many employers begin the hiring process with a written questionnaire, then conduct interviews with finalists before making decisions. Frost strongly endorsed this approach.
“A structured live interview is one of the most effective mitigation strategies,” he stated. “It quickly becomes obvious whether a candidate can explain or replicate earlier responses.”
Given that the line between polishing and embellishing is not always clear, Frost cautioned entrepreneurs not to overreact when evidence of a jobseeker using AI arises.
“The most effective response is to take a measured, context-driven approach rather than applying a one-size-fits-all rule,” he counseled. “The next steps depend on the severity and intent behind the behavior. For example, a highly capable, motivated candidate might have used AI tools to polish their assessment responses — not to misrepresent who they are, but to put their best foot forward. If that same candidate performs strongly in a live interview and their verbal responses align with the skills demonstrated in the assessment, there’s good reason to continue moving them through the process. Not every flagged behavior is deception; sometimes it’s just a nudge to dig a little deeper.”
While entrepreneurs certainly need to be mindful of the presence of AI in the hiring process, Frost currently sees no cause for alarm.
“With the right safeguards in place, organizations can confidently filter out the true bad actors,” he said, “while still giving great candidates a fair shot.”
Keith Twitchell spent 16 years running his own business before serving as president of the Committee for a Better New Orleans from 2004 through 2020. He has observed, supported and participated in entrepreneurial ventures at the street, neighborhood, nonprofit, micro- and macro-business levels.
