As the first substantive conversation between candidate and recruiter, pre-screens set the tone for the recruiting process. Increasingly, parts of this step are being delegated to AI, changing candidates’ first impressions and what they expect from the interactions that follow.
AI can speed up recruiting workflows by collecting and organizing information, saving recruiters time and helping strong candidates progress faster. When it takes over conversations that require judgment or trust, however, it can create friction and discourage candidates from continuing.
The boundary depends on the task being delegated. Collecting a candidate’s availability is different from interpreting their career history or understanding why they would leave a successful role.
What pre-screening is for
A pre-screen is a short call between a recruiter and a prospective candidate before an interview with the hiring team. It has two primary jobs.
- Checking facts: the recruiter confirms location, availability, compensation expectations, credentials, and the client’s hard requirements.
- Discovery: the recruiter explores why the candidate would consider a move, what makes the opportunity compelling, and what might stop them from proceeding. In senior and specialist searches, this includes distinguishing polite interest from a willingness to change jobs.
Both inform whether the candidate moves forward. The first lends itself to a checklist. The second requires trust, careful listening, and follow-up questions shaped by the conversation.
Discovery also helps candidates assess the opportunity. Recruiters need to explain why a move could make sense, answer questions, and understand what the candidate needs before committing to the process.
Three levels of AI in screening
AI can support several parts of screening, but each use case needs to be assessed by what the system is being asked to do.
Resume screening and filtering. AI can extract credentials, organize employment history, and highlight experience against stated requirements. Ranking candidates or rejecting them based on inferred fit goes further. Those tasks involve judgments about what experience matters and how different careers paths should be evaluated.
Narrow conversational screeners. A tool asks defined follow-up questions, such as whether a candidate holds a required license, can work specified shifts, or is willing to relocate. Its role can remain limited to collecting answers and flagging matters for recruiter review.
AI-led interviews and assessment. A conversational agent conducts an interview through text or voice, potentially adapting its questions and evaluating responses. This gives the system a broader role in interpreting the candidate.
One-way video interviews are a separate format. Candidates record answers to preset questions, which may be reviewed by people, scored by AI, or both. A recorded interview is not necessarily an AI interview, and its limitations should not be confused with those of a live conversational agent.
The wider adoption trend makes these distinctions increasingly important. SHRM’s 2025 Talent Trends data reports that organizational use of AI in HR tasks reached 43%, up from 26% in 2024. That measures HR adoption broadly, rather than adoption of any particular screening format.
Where AI helps screening
AI is most useful in screening when its task is clearly defined, such as collecting information or checking answers against the explicit requirements. It can organize responses for review and highlight missing details, giving recruiters more time to assess fit and explore motivation.
Consider event staffing. A recruiter filling concert shifts may need to confirm availability, location, required certifications, and willingness to accept the stated pay. Much of that information can be collected through a structured exchange.
A specialist search may also benefit from automated information gathering. Confirming a professional license or notice period can help a recruiter prepare for a more productive conversation.
The boundary appears when an answer needs interpretation. A compensation expectation above the stated range may warrant a discussion about the overall package. An unfamiliar job title may conceal highly relevant experience. Missing information may call for a question rather than a rejection.
Collecting an answer and deciding what it means are different responsibilities. Recruiters should remain accountable for the interpretation and the decision to advance a candidate.
What automation can take away
Problems start when the screening process asks candidates to share sensitive information or make consequential decisions without adequate access to a recruiter.
Candidate research illustrates the risk. In a recent Greenhouse survey of job seekers, 38% of U.S. respondents said they had walked away from a hiring process because it included an AI interview. The report also identified prerecorded video interviews scored by AI without a human present as a reason candidates might withdraw. These findings concern AI interviews and AI-scored recordings, rather than every use of automation in screening.
Several factors matter for recruiters:
- Candidates may need reassurance before answering. A financial professional may be reluctant to discuss client relationships, business performance, or information connected to their current employer. A recruiter needs to understand the concern, explain what is relevant, and respect appropriate limits. An incomplete answer should prompt clarification before it becomes a negative judgment.
- Candidates need to ask questions, too. A screening tool may answer routine questions about the role, but discussions about leadership expectations, organizational stability, or career progression often require someone with direct knowledge of the client.
- Big decisions require an ongoing conversation. Relocation, compensation changes, and leaving a long-standing employer can affect a candidate’s household and professional relationships. Recruiters help candidates work through those decisions over time.
- Trust cannot be assumed. Gartner‘s 2025 research found that only 26% of candidates trusted AI to evaluate them fairly, even though 52% believed it screened their application information.
These findings do not mean candidates reject every automated interaction. Greenhouse also found that some AI interview experiences improved candidates’ impressions of employers. Clear disclosure, an explanation of what is being assessed, and access to a human were prominent candidate preferences.
In senior and specialist searches, recruiters should consider whether the convenience gained from automation is worth the friction it may introduce with a prospective candidate.
The overlooked cost: how junior recruiters learn
In many recruiting teams, pre-screen calls have served as an informal training ground. Junior recruiters learn the vocabulary of their market, connect client requirements to actual career histories, and discover which follow-up questions reveal useful information.
Preparing for those calls matters, too. Reading client notes, breaking down a role, and building search criteria forces recruiters to understand the assignment before discussing it with candidates.
When AI handles the preparation and the first exchange, firms risk removing those learning opportunities without replacing them. AI can accelerate access to market knowledge, but firms still need to teach recruiters how to apply that knowledge in candidate conversations. Knowing what a credential means is different from understanding its relevance to a particular search.
Firms can preserve that development in three practical ways:
- Require preparation before reviewing AI output. Junior recruiters should explain the role’s essential requirements, likely candidate motivations, and questions they would ask.
- Bring junior recruiters into live conversations early. Shadowing experienced recruiters shows how to discuss sensitive moves, address concerns, and explain an opportunity credibly.
- Review automated screens together. Discuss what the answers establish, what remains uncertain, and which follow-up questions would resolve it.
Where the line sits
Automate clearly defined information gathering. Keep recruiters responsible for interpreting answers, exploring motivation, and helping candidates assess the opportunity.
That boundary applies across resumes, chatbots, recorded interviews, and conversational agents. The question is what responsibility the tool has been given and where recruiter judgment enters the process.
Firms that draw this line carefully can gain speed without weakening the conversations that determine whether a candidate proceeds or accepts an offer. They also need to train recruiters deliberately for that work as routine information gathering becomes increasingly automated.