Practice vs. the real thing: where AI help ends in interview prep and assessment
AI can help candidates prepare, but employers are drawing a sharper line around live performance, disclosure, and final judgment. The practical question is no longer whether to use AI at all, but when it stays in practice and when it must stop.
Where practice ends and performance begins
What is the line between using AI to prepare and using it in the moment? That question is getting more important as employers adopt AI in early screens and also set clearer rules about what candidates may do.
According to Greenhouse, candidates may use online research, their networks, recruiter questions, and AI to prepare for interviews. But they should not use AI to supplement performance in a live interview unless they are instructed or have agreed to do so.
The rules are becoming stricter, not looser
Greenhouse says candidates should not read from or reference AI-produced answers during interviews without permission or advance disclosure. It also says that using AI outside its guidelines, without disclosure or citation, can disqualify a candidacy.
That matters because the boundary is no longer vague. The company’s guidance separates preparation from performance and makes disclosure part of the rule, not an optional courtesy. If AI is part of the process, the candidate has to know whether it is allowed, how it should be used, and whether it needs to be named.
Assessments are different from interviews
The same distinction shows up in work assessments. Greenhouse says some assessments explicitly encourage AI use for research, iteration, or coding exercises, but the final submission should still reflect the candidate’s own synthesis, decisions, and point of view.
That is a useful standard because it does not ban tools outright. It treats AI as support for thinking, not a substitute for thinking. In other words, the question is not whether a candidate used AI somewhere along the way, but whether the work being judged is still the candidate’s own judgment.
AI in the hiring process still depends on humans
Greenhouse’s hiring process also includes a Voice AI stage for certain positions, which it describes as agentic interviewing to get a better signal of experience beyond the written resume. Greenhouse says its hiring teams define the questions and rubric, and humans evaluate the results before making a decision.
That human layer is not incidental. Greenhouse’s Notetaker terms say interview recordings, transcripts, notes, and summaries are assistive inputs, and that hiring evaluations and decisions must be made by human reviewers. The terms also say customers must not use the tool to score, rank, rate, recommend, or automatically assess candidates. OpenAI’s Help Center makes a similar point: it is inadvisable and against its Usage Policies to rely on models for assessment decision purposes without a human in the loop.
Why disclosure and trust now matter more
The candidate side of this is not theoretical. Greenhouse’s 2026 Candidate AI Interview Report says 63% of U.S. job seekers have already experienced an AI interview, and 70% of those who experienced AI evaluation were not clearly told AI was being used before their most recent AI interview. The same report says 38% of U.S. candidates have withdrawn from a hiring process because it included an AI interview, and another 12% would if required.
Those numbers show why the rules around AI assistance are becoming more explicit. Candidates are being asked to prepare with AI, but they are also being asked to trust systems that may be used before a human speaks to them. The practical response is straightforward: use AI for practice, read the employer’s or school’s rules carefully, disclose tools when required, and if an automated screen seems wrong, raise it through the process the organization provides.
Sources
- Guidelines for using AI in our interviewing process Greenhouse · September 30, 2026
- Greenhouse Notetaker Terms Greenhouse · July 31, 2026
- How can ChatGPT be used for assessment and feedback? OpenAI Help Center · September 28, 2026
- AI interviews in hiring: What candidates actually want – and how to get it right Greenhouse · April 29, 2026
This article was researched and drafted with AI from the sources above. Spot an error? Email info@hiddenproai.online.