The Person You Interviewed May Not Be the Person Who Shows Up.

The interview went well. That is no longer proof of who you interviewed

The case rarely announces itself. A candidate for a data role interviews by video over two or three rounds, answers the technical questions well, and accepts the offer. Then the problems start in small pieces. The person who arrives looks or sounds different from the person on the calls. The new hire cannot repeat work they explained fluently a month earlier. A background check detail does not line up with what the interview team was told. In the serious cases, which security researchers now document openly, the interview was taken by a stand in, a more qualified person hired to pass the process for someone else, or the identity itself was borrowed or invented so that a stranger could get inside a company's systems.

This is a different problem from the exaggerated resume, and it needs a different response. A candidate who inflates their skills is a quality problem, and live problem solving exposes it. A candidate who is not who they claim to be is a security problem. For most roles, a fraudulent hire wastes a salary and a quarter. For a data role, the hire is handed the keys on the first week: database permissions, reporting platforms, customer records, and the internal files that explain how the business runs. Gartner's researchers put the point plainly: candidate fraud creates cybersecurity risks that can be far more serious than making a bad hire. When the role touches data, verifying skills without verifying identity leaves the larger risk wide open.

What the evidence says

Checkr's second annual hiring fraud report, The Hiring Hoax 2026, is based on an online survey of 3,000 United States managers who had been involved in hiring within the previous 12 months, conducted September 11 to 16, 2026. Forty percent of those managers had interviewed a candidate who was later found to be using a fake identity, up from 31 percent in the 2025 report. Sixty five percent had uncovered candidates who misrepresented their skills or experience, up from 60 percent. Sixty percent said artificial intelligence has made virtual interviews harder to trust, up from 51 percent, and 68 percent believed candidates are now better at using AI to fake their identities than companies are at detecting them, up from 62 percent.

Candidates themselves confirm the behaviour is real, not just suspected. In a Gartner survey of 3,000 job candidates in the second quarter of 2025, 6 percent admitted to participating in interview fraud, either posing as someone else or having someone else pose as them in an interview. Gartner predicts that by 2028, one in four candidate profiles worldwide will be fake. That forecast is cross checked in reporting by HR Dive and in other publishers citing the same Gartner research, and the direction matches Checkr's year on year rise. Fraud of this kind does not need to be common to be costly. It needs to reach one data hire with broad access.

The confidence gap is the part a hiring manager can act on. Formal fraud training among HR teams rose sharply in Checkr's research, from 17 percent to 71 percent, and 62 percent of organizations now use third party background checks, up from 53 percent. Yet only 23 percent of managers were extremely confident their hiring process would catch a fraudulent candidate. Pressure explains part of the gap: 28 percent of managers admitted they had knowingly advanced a candidate despite suspecting dishonesty because they needed to fill the role quickly. The blind spot after hiring is wider still. Thirty eight percent of managers suspected a fraudulent employee was already working at their company undetected, while 63 percent of organizations focused fraud prevention mainly on the stage before hiring and only 6 percent prioritized prevention after hiring, which is exactly when system access is granted.

What serious firms verify, and when

The firms handling this well have not made hiring slower or more suspicious. They have moved verification to set points, applied the same steps to every candidate for a role, and written the steps down so no stage depends on one interviewer's memory or instinct.

1. 1State the rules at the start: tell every candidate what use of AI is acceptable in the process, that identity is verified, and that fraud ends the process, as Gartner recommends in its published guidance
2. 2Verify identity before the final round, not after the offer: government issued photo identification matched against the live person on camera, by a trained checker rather than a screenshot in an email thread
3. 3Keep the same person across every stage: interviewers note who they actually saw and heard, cameras stay on for technical rounds, and any change in appearance, voice, or ability between stages is treated as a stop sign, not an awkward detail
4. 4Test the work live and in person where the role allows it: a supervised work sample or a final on site round makes a stand in arrangement impractical, and Gartner found 62 percent of candidates said they are more likely to apply to a position that requires in person interviews
5. 5Confirm the record directly: employment and credentials checked with the issuing employer or institution, references reached through independently found contact details rather than numbers the candidate supplies
6. 6Verify again at onboarding, before access: identity rechecked on day one against the interview record, then access granted in stages, with production data and customer records released only after the hire is confirmed and settled

Two disciplines in that sequence matter more than the tools. First, consistency. The same checks, in the same order, for every candidate for the role protects the company twice over: it closes the gaps a fraudulent candidate looks for, and it treats honest candidates fairly, because no one is singled out on a hunch. Gartner makes the same point about trust, advising employers to pair fraud prevention with clear communication so legitimate candidates know what to expect and why. Second, no exceptions for speed. The 28 percent of managers in Checkr's survey who advanced a suspect candidate to fill a role quickly were not careless people. They were busy people with an open seat and a deadline. A written process removes that decision from the worst possible moment, which is the moment the vacancy hurts most.

What this means if you are hiring data talent in Oklahoma

Oklahoma firms hiring data professionals work in a small, connected market. Bureau of Labor Statistics figures put the state's information sector at about 16,700 jobs in September 2026, within a statewide unemployment rate of 4.2 percent. A market that size cuts both ways. Locally, reputations travel, and an in person final round is easy to arrange in Oklahoma City or Tulsa in a way a national remote employer cannot match. That is a genuine security advantage, and local firms should use it deliberately rather than copy an all remote process built for another market. At the same time, most Oklahoma companies now interview at least some data candidates remotely, including candidates from outside the state, and remote stages are exactly where identity fraud operates. One fraudulent hire also lands harder here. A lean Oklahoma data team has no spare capacity to absorb a quarter of bad access decisions, a repeated search, and the audit of everything the wrong person touched.

At Oklahoma Data Partners, identity and skills verification are built into our data recruiting work from the first conversation: candidates verified as the people they claim to be, skills proven with live work rather than a polished document, and the same person confirmed from first interview to first day, before any client system access is discussed. If you are hiring a data professional who will hold real access to your systems and your customers' information, book a consultation and we will show you where identity checks belong in your process and which stages of your current hiring would let a stand in through.