Combating “Skillfishing”
Both candidates and employers are increasingly turning to AI throughout the employment process and it's changing how each side has to operate. Candidates are using AI to help construct, and in some cases fabricate, resumes and cover letters, while employers lean on AI as a preliminary screening mechanism. One emerging tactic on the candidate side has been dubbed "skillfishing": using AI to pad an applicant's skillset and trick employers into considering them for roles they aren't actually qualified for.
The dynamic isn't new so much as it's accelerating. Employers have long embedded specific keywords in job postings to filter for applicants with the requisite skills, but job seekers have been wise to that strategy for years, salting their resumes with the same keywords to clear the initial screen. AI simply makes that keyword-matching far easier to do at scale. Some technologically advanced candidates have gone as far as building fake interviewee personas complete with the ability to fake video interviews.
How should employers approach this growing phenomenon? A management professor at George Mason University frames it as "hire hard, manage easy". Employers should invest real time upfront in a selection process that assesses skills, knowledge, and cultural fit, rather than trying to sort things out after a bad hire is already on the team. In practice, that means building a screening and interview process specifically designed to thwart skillfishing and AI-assisted deception.
Here are several tactics to consider:
Skill-based interrogation. More rigorous, pointed questioning during interviews forces applicants to reveal how much, or how little, they actually know about the skills the role requires.
Mini-cases and situational questions. For online screening, scenario-based prompts require candidates to demonstrate real skills earlier in the process, rather than just reciting keywords.
In-person interviews. These remove the opportunity to lean on AI for real-time answers (unless testing that capability is itself the point of the interview) and are especially useful for higher-level roles individually, or lower-level roles in a group format.
Open-ended and scenario questions. Questions that preclude simple yes/no answers are far better at producing an accurate read on a candidate and at defeating an automated or AI-coached response.
A consistent set of core questions. Recruiters recommend using the same base questions for all applicants to create a uniform basis for comparing candidates against both the position requirements and each other.
Probationary periods, temporary contracts, or contingency hires. These offer a real-world proving period, though they're not a perfect fix: the implicit "threat" built into a probationary arrangement can create its own problems. If used, employers should set clear expectations upfront and provide regular feedback and reasonable support for managing minor skills gaps or integration issues — taking the "long view" so an employee can't later claim they were blindsided.