Should We Stop Sending Interview Questions in Advance Since Candidates Are Running Them Through AI?

Key Takeaways

• Send only behavioral questions ahead for thoughtful responses
• Use deep follow‑up probes to verify candidate authenticity
• Treat AI‑generated answers as diagnostic data, not disqualifier
• Provide topic outlines for technical questions without exact wording
• Set clear policy; call out AI use during interview

Summary

Sending interview questions 30 minutes before a remote interview aims to boost equity and candidate comfort, but generative AI enables applicants to produce polished, pre‑written answers, compromising authenticity. Hiring panels worry that AI‑generated responses mask true competence, prompting a debate over whether to continue the practice. A balanced approach suggests limiting pre‑sent questions to behavioral prompts that require personal anecdotes, while keeping technical queries confidential. Follow‑up probing during the interview can further verify the candidate’s genuine expertise.

Pulse Analysis

The practice of sending interview questions to candidates minutes before a remote interview has grown as a diversity‑and‑inclusion measure, giving applicants time to prepare thoughtful answers and reducing anxiety. However, the rapid adoption of generative AI tools has introduced a new loophole: candidates can feed the questions into large‑language models, receive polished responses, and present them as their own. This undermines the original intent of equitable assessment and raises concerns for hiring managers who rely on genuine insight into a candidate’s problem‑solving and communication skills.

A pragmatic compromise is to limit pre‑sent material to behavioral prompts that require personal anecdotes, such as “describe a time you managed a difficult client.” These questions naturally demand specific experiences that are difficult for AI to fabricate convincingly. For technical or case‑study topics, share only high‑level themes or a brief agenda, keeping the exact phrasing confidential. During the interview, employ layered follow‑up questions—probing challenges, tools used, decision rationale—to test depth of knowledge and expose any reliance on canned AI output.

Beyond the immediate interview, AI‑generated answers can serve as a diagnostic signal. Candidates who attempt to pass off generic text reveal a willingness to shortcut critical thinking, a trait that may predict future performance issues. Organizations should codify a transparent policy that outlines acceptable preparation practices and empowers interviewers to call out suspicious AI use in real time. By balancing accessibility with rigorous probing, companies protect the integrity of their talent pipeline while still fostering an inclusive candidate experience in an AI‑augmented hiring landscape.