Fake candidates are now the number one hiring challenge of 2026. Not the talent shortage. Not comp. Fraud. GoodTime's 2026 Hiring Insights Report put AI-generated fraudulent candidates at the top of the list, above every other problem staffing leaders named. Gartner expects one in four candidate profiles to be fake by 2028.
That is not a future problem. It is already in your pipeline. Deepfake video, AI-written resumes, stolen identities, and proxy interviewers are cheap and easy to run at scale. Someone can sit in one interview and have a second person do the actual work. Someone can pass a video screen with a face that is not theirs.
The good news: you do not need to rebuild your tech stack to catch most of it. You need an audit and a protocol. Here is the one to run before Q4.
Why this hits staffing firms harder than corporate HR
A corporate recruiter fills a few roles a month. You fill hundreds. Your volume is the whole point of your business, and it is also the attack surface. The more reqs you move, the more chances a fraudster has to slip through.
Then there is speed. Your recruiters get paid to fill fast. Fraudsters know this. They target high-volume, remote, and contract roles because those move quickest and get the least scrutiny. IT contract work is the top target right now. So is anything fully remote.
And the cost lands on you twice. First when you place the fake candidate. Second when your client finds out and questions every other placement you sent them. One caught fraud can put a whole account under review.
The 5-step fraud audit to run this quarter
This audit takes a focused week. Assign one person to own it. Do not spread it across the team or it will never finish.
Step 1: Pull your fast-fail data
Export every placement from the last six months that ended inside 30 days. For each one, note the reason. You are looking for day-one no-shows, performance that collapsed immediately, or a client saying the person who showed up was not the person they interviewed. Count them. This is your baseline fraud rate, and most firms have never measured it.
Step 2: Audit your identity checkpoint
Find the exact moment in your process where you confirm a candidate is who they say they are. If you cannot name that moment, that is the finding. Most firms verify identity at onboarding, which is far too late. By then you have already sold the client on someone you never confirmed was real.
The fix is a live, on-camera ID check early in the process. Camera on, government ID held up, face matches. It takes 60 seconds. Do it before the client interview, not after the offer.
Step 3: Test your remote interview for proxy attacks
A proxy attack is when one person interviews and a different person does the job. To catch it, add one unscripted question that forces a real-time answer about the candidate's actual experience. Not "tell me about yourself." Something like "walk me through the last bug you fixed and what broke first." A coached proxy stumbles on specifics. A real practitioner does not.
Also watch for signs of AI assistance during the call: eyes tracking a second screen, a slight delay before every answer, audio that does not match lip movement. Train your recruiters to notice these and flag them.
Step 4: Run a resume and profile sanity check
AI writes clean resumes. Too clean. Look for perfect keyword matches to your job description, employment dates with no gaps ever, and a LinkedIn profile created in the last 90 days with few connections. None of these alone proves fraud. Two or three together earns a closer look.
Add a reverse image search on the profile photo. Stolen and AI-generated headshots turn up fast. This costs nothing and takes a minute.
Step 5: Check your reference process for holes
Fraudsters supply fake references who are in on it. Stop calling numbers the candidate gives you. Find the reference independently through the company's main line or a verified profile. If a reference only exists on a phone number the candidate provided, treat it as unverified.
Turn the audit into a standing protocol
An audit you run once does nothing. The point is to build a protocol your team runs on every candidate without thinking about it.
Keep it tight. Three checks that happen every time beat ten checks nobody follows. Here is the minimum:
- Live ID check on camera before the client interview.
- One unscripted skills question in every screening call.
- Independent reference contact, never a candidate-supplied number.
Write these into your recruiter checklist. Make them required fields in your ATS so a req cannot advance without them logged. That last part matters. If it is optional, it will not happen under deadline pressure.
Track your fast-fail rate every month after that. If your protocol works, the number drops. If it does not drop, your protocol has a gap and you go find it. This is how you keep fraud from creeping back in as the tactics evolve.
One more thing for the PE portfolio leaders reading this. Run this audit across every firm in your portfolio, not just the flagship. Fraud rates vary wildly between offices, and the weakest one drags down the brand for everyone. A standard protocol across the group is worth building now.
Do this week
Pick one person and have them pull the fast-fail data from Step 1 by Friday. That single report tells you whether you have a small problem or a big one, and it takes a few hours. You cannot fix what you have not measured, and right now most firms are flying blind on this. Get the number first. Then build the protocol around it.