# LinkedIn's 6-Minute AI Screen Checks Your Resume. Say the Same Numbers

_LinkedIn's Hiring Assistant 2 runs six-minute voice pre-screens and puts your answers next to your resume. Here's what it asks and how to answer consistently._

- Canonical URL: https://www.refolk.ai/candidates/blog/linkedin-hiring-assistant-2-voice-pre-screen-resume-consistency
- Published: 2026-11-08
- Reading time: 9 min
- Keywords: LinkedIn Hiring Assistant pre-screening questions, LinkedIn AI voice screening interview, how to answer AI pre-screen questions, LinkedIn Hiring Assistant 2 candidates, AI phone screen tips

If you just got a LinkedIn pre-screen invite, how you sound matters less than whether your answers match your resume. LinkedIn's Hiring Assistant 2 is free for English-language customers from November 2026. It runs pre-screens, now including voice, with a median length of six minutes. Then it shows what you said next to your resume, your work history and your verified employers, so it can flag inconsistencies. Awkward candidates aren't the ones who get hurt. The ones who get hurt are candidates whose numbers change between the page and the call.

## What is LinkedIn Hiring Assistant 2?

Hiring Assistant 2 is LinkedIn's recruiting agent. It's software that pre-screens, evaluates and sorts applicants for a recruiter, and it can now run a short voice interview too. LinkedIn announced it on September 29, 2026 at Talent Connect in New York (one outlet dates it October 1). The post came from Dan Reid, VP of Product, and Prashanthi Padmanabhan, VP of Engineering. It's free for English-language customers from November 2026.

The scale is the reason to prepare. More than 20,000 organizations already use LinkedIn's agentic hiring tools. With that footprint, a pre-screen may well be your first conversation with an employer, not an optional extra.

> **6 min** Median time for Hiring Assistant to complete a pre-screen
>
> LinkedIn compares it with the 3.5 hours a candidate typically waits for a recruiter's first response.

Recruiters are buying speed. What that means for you: your first "interview" happens before any person has read your file, and it produces a structured record that a person will read later. That record is what you're preparing for.

## What LinkedIn Hiring Assistant pre-screening questions ask

The pre-screen asks basic questions in three areas: competency, work authorization and location, and it can now ask them by voice. None of them is hard, but each one is a place where what you say can contradict what you wrote.

- **Competency.** Expect questions like "How many years have you worked with Python?" or "Have you managed a team?" The risk is rounding up. If your dated roles add up to three years and you say "about five," that becomes a visible gap.
- **Work authorization.** "Are you authorized to work in this country without sponsorship?" The risk is a casual spoken answer that differs from the box you ticked on the application form.
- **Location.** "Where are you based, and are you open to relocating?" The risk is saying one city when your profile and resume list another.

Basic questions produce answers that can be checked against a structured field. Nobody can cross-check your answer to "Tell me about a project you're proud of." Anyone can cross-check "years of Python" against the dates on your resume.

## What the recruiter sees after the call

The recruiter sees a candidate card. It shows your identity verification, verified employers, pre-screening responses and work history side by side, along with the agent's own assessment. According to ERE's reporting, the agent "will combine these validations with ATS data, pre-screening answers and work history to provide a richer portrait of the candidate and flag inconsistencies."

That last phrase explains the whole design. LinkedIn built this to deal with fraud:

- **39% of recruiters** name not knowing whether an applicant, or their skills, is real as a top challenge.
- **115 million members** are verified on LinkedIn, which gives the agent an identity layer to check against.
- **6% of candidates** in a 2Q25 Gartner survey of 3,000 job seekers admitted to taking part in interview fraud. Gartner HR senior research director Jamie Kohn has been quoted on the problem.

The mechanism matters. A system built to catch fake candidates treats any mismatch as a signal, whatever the intent behind it. Say an honest candidate rounds three dated years up to "about five." On the card, that looks exactly like a padded claim. The agent can't tell sloppiness from deception, and a busy recruiter scanning a stack of cards won't stop to work it out either.

> A nervous candidate with consistent facts beats a polished one whose numbers drift between the page and the call.

## Applying outside LinkedIn doesn't take you out of the check

If the employer uses one of the connected ATSs, an application you sent outside LinkedIn goes into the same evaluation. An ATS (applicant tracking system) is the software employers use to collect and manage applications. It's what sits behind the "Apply" button on a company's careers page.

LinkedIn publishes two separate lists, and they're easy to mix up:

| Capability | Systems named |
|---|---|
| Evaluate applicants who applied off LinkedIn | Greenhouse, Lever, SmartRecruiters, Zoho Recruit, Tracker |
| Evaluation "coming soon" | iCIMS, Workday |
| Show verification badges | iCIMS, JazzHR, Jobvite, Lever, SmartRecruiters, Workable |

Lever and SmartRecruiters appear on both lists. Greenhouse can feed your application into the evaluation but isn't on the badge list. Greenhouse's own help docs say connected applicants are evaluated using resume data. So the resume you uploaded to a Greenhouse careers page is what the agent reads, not your LinkedIn profile alone.

Here's what that means in practice. The resume you tailored for this job, the LinkedIn profile you last updated two years ago and the answers you give on the call can all end up on one card. Tailoring is still the right move, but it should change what you **emphasize**, never the **facts**. Retitling "Senior Analyst" as "Analytics Lead" on one version, or stretching a contract into a full-time role, now leaves a visible trail.

This is the exact problem [Refolk](/candidates) is built around. It writes every tailored resume from one record of your own history. Each version reorders and reframes for the posting, but the titles, dates and numbers come from the same source, so the copy in Greenhouse and the profile on LinkedIn agree.

## Recruiters barely see the plumbing, so assume it's connected

Assume your careers-page application feeds LinkedIn's agent, because nobody is likely to tell you it does. Refolk's index of professional profiles suggests recruiters treat the ATS as background plumbing, even though they use it every day:

| Segment (Recruiter, Technical Recruiter, TA Specialist, TA Partner) | Count |
|---|---|
| US recruiters and talent acquisition | 145,051 |
| UK recruiters and talent acquisition | 11,678 |
| US recruiters listing Greenhouse as a skill | 292 |
| US-to-UK multiple | 12.4x |
| Share of US recruiters listing Greenhouse | 0.20% |

These are the people most likely to read your pre-screen, and the most common title in the US sample was Technical Recruiter. The 292 who list Greenhouse work at employers including Meta, Snowflake, Warby Parker and Datavant. The UK figure probably understates the market, because the common British title "Recruitment Consultant" wasn't part of the query.

The 0.20% figure is the useful one. Greenhouse is one of the five systems feeding Hiring Assistant 2, yet about 1 in 500 US recruiters lists it as a skill. If recruiters don't think about the pipe, they won't think to warn candidates about it. That's why consistency is your job, not something the process will prompt you to check.

## The agent sorts, it doesn't reject. The sort still decides.

LinkedIn says Hiring Assistant does not screen out applicants or make hiring decisions on its own. That's accurate, and it leaves something out. Inside Greenhouse, recruiters can filter applicants into "Top Fit," "Might be a fit" and "Might not be a fit" based on the agent's evaluation. In practice, recruiters work from those fit buckets, and the top one gets read first.

> **4x** How much more likely recruiters are to contact candidates the AI suggests
>
> LinkedIn's own figure, which makes the bottom bucket function almost like a rejection.

No person rejected you, so no rejection email arrives. Your application just sits under a filter nobody opens. From your side, that's the familiar experience of applying into silence. You keep sending applications, the pre-screens finish in six minutes, and then nothing happens. A mismatch on the card is one of the quiet reasons that happens.

The way out of the bottom bucket is mostly earlier than the call. Apply where your history genuinely fits the posting, and make sure every version of that history says the same thing, so the agent has less to doubt.

## How to answer AI pre-screen questions without contradicting yourself

Answer every question with the same figures, titles and dates that are on the resume this employer has, and say them the way they're written. Do this before the call:

1. **Build a fact sheet from the resume you submitted, not from memory.** List each employer, title, and start and end month. If you tailored the application, read from the version that went to this posting.
2. **Do the years math for each skill.** Count only the roles where your resume shows that skill. If it adds up to just over three years, say "a little over three years," not "about five."
3. **Pick your headline numbers and say them word for word.** If the resume gives a percentage, say that percentage, not "roughly halved it."
4. **Write your work authorization answer as one sentence** that matches the application form exactly, including any end date or sponsorship need.
5. **State your location the way your profile does.** If you're relocating, say where you're based now and where and when you're moving, rather than naming only the destination.
6. **Reconcile your LinkedIn profile before the call.** Titles and dates should match the submitted resume. If one is out of date, fix it now, not after the card is generated.
7. **When a fact genuinely differs, give the reason in the answer.** "My title was Analyst, but I led the team for the final year" turns a flagged mismatch into context the recruiter can read.

The common thread: don't improvise any number during a pre-screen. Every figure you say out loud should already be in writing somewhere the agent can see.

## AI phone screen tips: prepare the content, not your voice

Put your prep into what you say, not how you sound, because tone shouldn't be scored. Emotion recognition in recruitment has been banned in the EU since February 2025, and vendor guides tell employers to switch tone scoring off. The pre-screen's stated scope is competency, work authorization and location, which are all content.

Candidates are right to be wary. In the same Gartner release, only 26% of job candidates said they trust AI to evaluate them fairly. You're entitled to ask the recruiter two plain questions: whether tone or delivery is scored, and how the pre-screen answers are used. A good recruiter will answer them.

A few delivery habits still help, mainly because they protect your content:

- **Lead with the number, then the context.** "Three years. Two at my previous employer, one in my current role."
- **Keep answers short.** Every extra sentence adds a claim that can be checked.
- **Don't fill silence with hedges.** "I mean, it's probably closer to five if you count side projects" creates a new figure that isn't on your resume.
- **Take the call somewhere quiet** so your answers come through exactly as you said them.

## FAQ

### Does LinkedIn's Hiring Assistant reject candidates automatically?

No. LinkedIn says Hiring Assistant does not screen out applicants or make hiring decisions on its own. It sorts them. In Greenhouse, recruiters can filter results into "Top Fit," "Might be a fit" and "Might not be a fit," and LinkedIn reports that recruiters are four times more likely to contact AI-suggested candidates. Landing in the bottom bucket rarely produces a rejection email, but it often means nobody reaches out.

### I applied on the company's careers site, not LinkedIn. Does the pre-screen still apply?

It can. If the employer uses Greenhouse, Lever, SmartRecruiters, Zoho Recruit or Tracker, applicants who applied off LinkedIn can be evaluated by Hiring Assistant 2, and iCIMS and Workday are listed as coming soon. Greenhouse evaluates connected applicants using resume data, so the resume you uploaded is what gets compared with your answers.

### Is the LinkedIn AI voice screening interview judging my tone or accent?

It shouldn't be. Emotion recognition in recruitment has been banned in the EU since February 2025, and vendor guidance tells employers to disable tone scoring. The pre-screen is built around competency, work authorization and location. If you're unsure, ask the recruiter directly whether delivery is scored, and spend your prep time on making sure your answers match your resume.

---

_From the Refolk job search blog. More posts: https://www.refolk.ai/candidates/blog/index.md_