7 Ways to Spot Fake Candidates (Without Scaring Off the Real Ones)

If “Morgan Freeman” in your DMs is funny, “Morgan Freeman” in your pipeline is where the joke ends. Fraudulent candidates, interview proxies, and deepfakes are here. And they’re here to stay.

Today, more than half of employers say finding qualified candidates is their number one challenge, with up to 23% of applications for high-volume roles showing evidence of being machine-generated.

Amid thousands of bogus resumes, real talent gets buried. To silence the bots and bring genuine candidates forward, you need to act fast on fraud and fair with people.

In this article, we’ll share seven signal‑first checks to help you catch fake candidates without scaring off qualified humans.

Finding genuine candidates in a sea of fakes

AI‑written resumes, interview proxies, North Korean scam rings. The signal‑to‑noise ratio is getting worse by the week.

Here’s what’s happening on the ground:

• Rising “resume slop.” AI‑generated applications flood pipelines, making signal‑first screening a must for remote and technical roles.

• Profiles don’t add up. Strange logos, odd career hops, blatantly mismatched timelines and locations.

• Suspicious LinkedIn. Broken or 404 profiles, stock headshots, 17 connections for “senior” candidates—the list goes on.
• “Proxy” interviews. Glitchy connections, reading answers off-screen, and making obvious slip-ups (like saying they’re from Denver but answering “Colorado” when asked which neighborhood.🙄)

With deepfake technology,  the sky’s the limit for career catfishing. But why would anyone pay proxies hundreds of dollars per interview just to land a job? 

Scammers have a range of motivations: steal your data, extort payouts, or use your brand as phishing bait aimed at employees and customers. 

Different motives, same mess for your team.

And recruiters aren’t the only ones on high alert. From 2020 to 2024, reports of job scams and losses to fake employment agencies nearly tripled to an eyewatering $501 million. 

Job seekers have their own list of red flags. Like recruiters, they’re using AI to help fight the mess it’s created.

To find qualified humans amid a sea of scammers, it’s time to refocus your recruiting.

7 ways to spot fake candidates

As an HR professional, it’s your job to protect the company. But candidates shouldn’t need two forms of ID, three references, and dental records just to score a phone interview.

To fight the fakes without alienating the real ones, you need small, consistent checks at just the right moment. Use these steps as your baseline, then test and iterate until you find what works.

1. Flag (but don’t eliminate) AI-generated cover letters and resumes

Tired of seeing “ChatGPT says” peppered into cover letters? You’re not alone. With 62% of hiring managers rejecting AI‑generated applications, authenticity checks are table stakes.

But don’t make the mistake of weeding out genuine candidates just because they use AI.

Instead, look for details that signal genuine interest:

• Clear impact stories. Look for clear role summaries, responsibilities, and metrics that show results. Not just a stream of logos.
• Education that fits the narrative. Dates align with credentials, with pivots explained in a sentence.
• Authentic look and feel. Templates are fine, but generic copy‑pasting and stock headshots are bigger red flags than ever.

These seemingly small authenticity issues can eat up a ton of time. And that’s where applicant tracking tools designed to catch AI-generated and invisible text can be a big help. 

According to a survey by ServiceNow, most candidates actually don't mind if you use AI in the hiring process, as long as it’s used as a support tool—not a replacement for hiring decisions.

Lead by example. Let applicants know upfront how much “AI polish” is too much. A simple blurb added your job description can get the job done:

“AI‑friendly, authenticity‑first. Use tools for polish, not to copy our JD or submit work you can’t explain. We run light fraud checks (JD copy/paste, timing, proxy) with human review; unverifiable applications may be declined.”

2. Raise the bar with prescreen questions

Candidate questionnaires add just enough friction to deter spray‑and‑pray applicants, while giving serious contenders a way to stand out. Keep questions short, role‑specific, and tied to the outcomes you actually care about.

Start with a minimalistic approach:

• Add 2–3 role‑relevant questions. E.g., “Tell us about a change you shipped that moved a key metric. What was the impact and how did you measure it?”
• Limit answers to 2–4 sentences to keep it scannable.
• Disqualify fully copy-pasted responses (or generic “enthusiast” statements with no substance. 🙄)

Start with the folks who spent five minutes on the questions. This small step helps genuine candidates rise to the top, instead of burying your team under manufactured CVs.

3. Ask for role‑specific work samples (plus context)

Credentials are easy to fake. In a market where 44% of job seekers say they would use AI to exaggerate or misrepresent their qualifications to an interview, work samples with context are your best antidote. 

Ask applicants for real work examples, plus a 2-minute video message with a screenshare explaining their approach. Once AI has done the initial step of prioritizing authentic candidates, the walkthrough can help you spot any deepfakes that may have slipped in under the radar.

Depending on the role, you can request samples like:

• For engineers: Recent repos or design docs, with commentary
• For product/design: Case studies with problem, process, constraints, outcomes, and visuals
• For data: Notebooks, dashboards, or write‑ups with dataset, methods, and business impact

Add a “process snapshot” in the job posting with information on how the review works, what you’ll look for (impact, narrative, decisions), and expected prep time (e.g., 20 minutes).

Create an automated “portfolio expectations” email when candidates apply thanking them for their application, recapping the requirements, and explaining how long the review takes. 

Bonus points if you can add an exact date when candidates can expect a reply.

⚙️ Systems tip: In Breezy, you can use stage actions to: 1.) auto‑send your verification policy before the call, 2.) capture quick notes in the interview scorecard, and 3.) advance verified candidates without extra unnecessary steps.

4. Run a real‑time skills check

Filters and voice changers and avatars. Oh. My.

For certain roles, security is a big concern. When the risk of hiring the wrong person is high, consider diving deeper via a live skills check or technical interview to verify identity in real‑time. 

Do your best to keep it candidate‑friendly:

• Use small, practical tasks. For example, edit a short code snippet or triage a bug.

• Ask for a brief screen share. Always ask for consent and provide alternatives if screen sharing isn’t an option.

• Encourage “think‑aloud” problem‑solving. Add a little ambiguity and watch how they clarify.

If a candidate can respond and adapt in real-time, you’re probably looking at a human. If you see glitches, airbrushed filters, or a Sora watermark—you’ve been deepfaked.

✅ Candidate experience check: Recent data shows that 83% of candidates will reject an AI interview. Consider using a video walkthrough (real-time or async) as an alternative, then invite your best aligned candidates to a live interview with a real person.

5. Verify during the live interview

If you’ve been burnt by deepfakes in the past, don’t beat yourself up. Even companies that specialize in catching phonies have wound up in interviews with bots. The technology is just that good.

Here’s a simple check you can run in the first 2–3 minutes:

• Start with clarity and consent. Let candidates know you’ll do a quick, two-minute identity check for security and fraud prevention.
• Run a liveness and location check. There are plenty of tools and APIs that can do this for you, or you can ask simple questions about the local weather, commute landmarks, or neighborhoods.
• Dig deeper with a situational question. Listen for strange lulls, pauses, or typing. Monitor  gestures closely and look out for any glitches in lighting, background, or facial expression.

For an added layer, you can ask the candidate to make a simple gesture like turning their head. If you notice filters, desync, or voice‑changers, it’s an instant red flag. Note the inconsistencies in your ATS so you and the rest of the hiring team can decide what to do next.

Don’t let one deepfake destroy the candidate experience for everyone else. 

As a general rule, use AI and automated tools like you want your candidates to. Keep it in a supporting role to help with auditing, sorting, and summarizing—and always keep the identity confirm human‑led.

💎 Keep it transparent: Make sure the job post includes clear expectations and an accurate salary range so candidates know their efforts are worthwhile. Publish a brief “what we verify and why” blurb, communicating why and when you collect this info—and how you protect it.

6. Add an in-person touchpoint for finalists

A brief onsite interview with finalists strengthens identity verification and team fit—without dragging out the process. Think short and purposeful: final interview, light-lift work task, or equipment pickup.

Here are a few other ideas:

• Add a short onsite interview, meet‑and‑greet, or equipment pickup before day one.
• Offer travel support where possible.
• Focus on identity and fit, not marathon interviews.

Even a single onsite touchpoint dramatically reduces proxy and deepfake risk while keeping things candidate‑friendly.

Set expectations in the job post so candidates know there’s one brief onsite step—even for remote roles.

7. Run the background check at offer stage

Save the heavier checks for when you’re really serious. Holding the final background check until just before onboarding helps keep the process quick, while protecting top candidates from over‑screening.

What to verify:

• Auto‑verify references and follow up with third‑party employment verification for titles, dates, and locations.
• Confirm role-relevant education, especially when timelines look off.
• Align checks with local laws and document your process.

Above all, keep the experience respectful. Check only what’s necessary and offer candidates a chance to explain any discrepancies.

Light checks, cleaner signals, better hires

In a chaotic hiring climate, your ATS should reduce noise—not add to it. Use your applicant screening tools to verify candidates quickly, without burying your team in admin.

Breezy Intelligence makes it easy:

• Signal‑first screening: Use Resume Audit to flag copy/pasted job descriptions and resume slop.
• Auto‑send questionnaires: To cut through the noise and surface thoughtful responses fast.
• Faster triage: Applicant Insights rank candidates on skills, experience, and overall fit.
• One workflow: Screening, outreach, scheduling, and scorecards all in one place.

Separate signal from slop, without slowing down. Breezy helps you build light checks into every stage so the best rise to the top. Give it a try with a free 14‑day trial.

FAQs

What are the biggest red flags for fake candidates?

Look for mismatches between the candidate’s resume and LinkedIn profile (job titles, years of experience, locations), plus AI‑generated cover letters, recycled resume designs, and thin social media footprints. In video interviews, watch for delays, scripted answers, crowded background chatter, and evasive responses to local or role‑specific questions.

Can recruiters rely on a LinkedIn profile to verify identity?

No. A LinkedIn profile proves an account exists—not authenticity. Use multi‑point checks: cross‑reference work history, job applications, and contact information; ask real‑time local questions; and pair with identity verification and a background check at offer stage.

How should hiring managers structure the interview process to catch fake applicants?

Keep it lightweight but real: prescreen questions tied to the job description, a 20‑minute “speak geek” deep‑dive, and short technical interviews that test problem‑solving in real time. Add one in‑person touchpoint for finalists—especially for remote jobs.

What’s the best way to verify location and identity without scaring off genuine candidates?

Use casual local checks during a video call (weather, neighborhood landmarks), confirm phone number ownership, and—when appropriate—request a redacted ID (name/photo only). Explain why you’re collecting it and how you protect data (fraud prevention + cybersecurity).

How do we spot AI‑generated resumes and bots at apply time?

Flag copy‑paste job descriptions, oddly perfect phrasing, and timing anomalies (resume created minutes before submission). Require 2–3 role‑relevant prescreen answers and weigh them in your screening process. Use AI tools in your ATS to automate fraud detection and route risky profiles for review.