Hired a Stranger? Where AI-Era Candidate Fraud Slips Through Your Hiring Process
Almost every founder we work with has a version of the same story. The candidate interviewed brilliantly. The CV was spotless. Three months in, something didn’t add up: a previous employer who had no record of them, work that didn’t match the skills shown in the interview, or a new hire who seemed oddly different from the person on the video call.
Those stories used to be rare. They aren’t any more, and the reason isn’t that candidates have become less honest. It’s that faking has become much cheaper.
What has changed
Three findings show the shift.
- Interview fraud is no longer fringe. In a 2025 Gartner survey of 3,000 job candidates, 6% admitted to interview fraud: either posing as someone else or having someone pose as them. Gartner predicts that by 2028, one in four candidate profiles worldwide will be fake.
- In India, the employment history is where the problems surface. EY’s 2025 study of more than one million pre-employment checks across 90+ Indian organisations found that 85% of discrepant IT/ITeS profiles were flagged on employment checks. And 45% of the discrepant IT/ITeS profiles were moonlighting: holding a second job, or running an active GST-registered business linked to their PAN.
- Recruiters are noticing. In a 2026 survey of 646 US HR professionals and recruiters by our BGV partner cFIRST, 98% said they had personally caught a candidate misrepresenting themselves, and 84% said it is more common than three years ago.
The pattern behind all three: the interview has become the easiest step to game, and the problems show up only later, in employment history, identity and day-to-day performance.
The new shapes of candidate fraud
| What it looks like | How it works | Where it usually surfaces |
|---|---|---|
| The AI-polished profile | CV, cover letter and LinkedIn profile written to match the job description almost word for word | Interview depth, employment verification |
| The coached interview | Real-time AI tools or a hidden helper feeding answers during a video interview | Follow-up questions, unscripted reasoning |
| The proxy interviewer | A skilled stand-in sits the interview; a different person joins | Joining day, first weeks of work |
| The identity switch | Documents belong to one person, the work is done by another | ID matching at joining |
| The double job | The new hire is still employed elsewhere, or runs a parallel business | UAN service history, performance, availability |
| The outsourced job | The hire quietly subcontracts their own work to someone cheaper | Output quality, system access patterns |
None of these is caught by a better job description or a longer interview. They are caught by checking the right thing at the right stage.
Where it slips through: stage by stage
Most growing companies run some checks. The gaps come from when they run them, and from what each stage assumes the previous stage already confirmed.
1. Application and screening
The gap: A CV that mirrors your job description gets shortlisted because it scores well, not because it’s true.
Close it:
- Treat a perfect keyword match as a reason to probe, not to fast-track.
- In the screening call, ask about one specific project: the team size, what went wrong, what they would do differently. Real experience produces detail; generated experience produces generalities.
2. The interview
The gap: Video interviews make it easy for someone else to answer, or to read answers fed in real time.
Close it:
- Keep cameras on, with no virtual backgrounds, for every round.
- Ask follow-ups that build on the candidate’s own earlier answer. Scripted help struggles when the next question depends on the last one.
- Hold at least one round live: in person, or with a screen share where the candidate works through a problem in front of you. Gartner found 62% of candidates are more likely to apply when an organisation requires in-person interviews, so this need not cost you good applicants.
- Use the same questions and scoring for every candidate. Inconsistency is easier to spot when everyone is measured the same way. (Our guide to structured interviews covers how to design them.)
3. The offer
The gap: The offer goes out before anyone has confirmed the employment history the candidate was hired for.
Close it:
- Make the offer conditional on verification, and say so in the offer letter.
- Start employment and education verification the day the offer is accepted, not after joining.
- Check UAN service history for overlaps between employers. Overlapping PF contribution periods are a clear sign of dual employment.
4. Joining day
The gap: The person who walks in is assumed to be the person who interviewed.
Close it:
- Match original ID documents to the person physically present.
- Have the hiring manager, not only HR, meet the new joiner on day one. They are the one who will recognise a different face or voice.
- For remote joiners, run a short live video call with the original interviewer during the first week.
5. The first 90 days
The gap: Probation is treated as a formality, so performance that doesn’t match the interview is explained away.
Close it:
- Set clear 30, 60 and 90-day goals and review them properly.
- Ask managers one direct question at each review: does this person’s work match what we saw in the interview?
- Make your exclusivity and conflict-of-interest expectations explicit in the employment agreement. (See key clauses every employer must include.)
Match the effort to the risk
A 20-person company can’t run every control on every hire, and doesn’t need to. Tier by what the role can access.
| Role risk | Examples | Minimum controls |
|---|---|---|
| Standard | Most individual-contributor roles with limited system access | Cameras-on interviews, conditional offer, ID matched at joining, employment verification |
| Elevated | Roles with client data, finance, code repositories or admin access | All of the above, plus one live or in-person round, education verification and UAN overlap check before access is granted |
| High | Finance heads, roles handling cash or sensitive personal data, roles working with vulnerable people | All of the above, plus court and criminal record checks and a closer 90-day review |
The single most useful rule across every tier: verify before access begins, not after.
Staying fair to honest candidates
Fraud controls should never make the process feel hostile. Most candidates are genuine, and they notice how they’re treated.
- Say it upfront. Tell candidates which checks you run and what use of AI tools is acceptable (polishing a CV is fine; being fed answers in an interview is not). Gartner recommends exactly this: clear expectations reduce fraud and build trust.
- Get consent properly. Verification involves personal data, so consent and careful handling are not optional. (Our guide to data privacy in HR under the DPDP Act explains what that means in practice.)
- Treat red flags as questions, not verdicts. A gap in employment history usually has an ordinary explanation. Ask before you assume.
A 30-minute check of your own process
Sit down with whoever runs hiring and answer six questions:
- Do we hold at least one live or in-person round for roles with sensitive access?
- Do our interviewers ask follow-ups that build on the candidate’s own answers?
- Is every offer conditional on verification, and does verification start at offer acceptance?
- Do we check UAN history for overlapping employment?
- Does the hiring manager meet every new joiner in person or on live video in week one?
- Does anyone get system, client or finance access before verification is complete?
Any “no” is a gap worth closing before your next hire. For a deeper look at the checks themselves and how to handle discrepancies, see our guide to background verification best practices for Indian employers.
Need a second pair of eyes on your hiring checks?
If you’d like a second pair of eyes on any of these gaps, reach out to us on our Let's Talk page (https://www.humanetics.in/contact) to review your hiring and verification process.
Sources
- Gartner, Gartner Survey Shows Just 26% of Job Applicants Trust AI Will Fairly Evaluate Them, press release, 31 July 2025: gartner.com
- EY India, The First Firewall: Background Checks as India Inc.’s Frontline Defense, May 2025 (data January 2024 to January 2025): ey.com (PDF)
- cFIRST, 2026 survey of 646 U.S. HR professionals and recruiters, as published by cFIRST Background Checks LLP on LinkedIn, October 2026.