With thousands of people applying for tech roles every month, the sheer scale of applications and candidates allows many fraudsters to go undetected. In fact, Gartner revealed that, worldwide, 1 in 4 job candidates will be fake by 2028.
HR systems now incorporate AI agents to help employers detect fraud. Fake candidates that companies are inadvertently hiring present fake résumés, scam websites, professional headshots, and LinkedIn profiles, CBS News reports. Fraudsters, once hired, can snatch company secrets or install malware.
Cliff Jurkiewicz, vice president of global strategy at Phenom, an applied AI company that specializes in HR, says that of the tech candidates on its platform, 40 percent to 60 percent of job applicants for its 700 IT clients are fake.
“It could take as much as three to 12 months to figure out that the person isn't qualified, especially once they get in the door and they're working remotely,” Jurkiewicz says. “They're going to use tools like the GenAI tools, especially if they're an engineer, to augment what they may or may not know, and they could get away with it for a little while.”
How Fake Candidates Operate
AI is allowing fraudsters to present fake credentials, and deepfake interviews are now popping up, says Abhishek Karnik, head of online threat research at McAfee.
Authentic candidates face the risk of being impersonated by hackers. In one notable instance, North Korean operatives posed as remote tech workers and secured employment at major U.S. companies. “Scammers often pull names, photos, and work history from real profiles to build fake personas,” Karnik says. “Even if you're not aware of it, your information could be part of someone else’s attempt to get hired under false pretenses.”
Fraudulent candidates get caught when they use examples during interviews of companies that do not exist, or they lack the same level of “authenticity” as a legitimate candidate, according to Jurkiewicz.
“You can game the system pretty easily by using a tool to make the perfect résumé,” Jurkiewicz explains. “Because recruiters are having to sift through [them], the problem is scale, so many that they're just looking at who is the best fit based on the recommendation of artificial intelligence. If they're not trained well on how to spot fraudsters, someone is going to make it to the next step and the next step and the next step if they continue to be able to use these tools.”
Take the case of an engineer hired by Jurkiewicz’s former software development company. The team noticed this engineer would farm out work to their family members in different parts of the world. As his team was auditing the candidate’s output, they also found junk in his code.
“That happens a lot, but this was particularly troublesome,” Jurkiewicz recalls. “We went in and started auditing the code, and we could tell, based on the project, that the same person wasn't writing the same code,” Jurkiewicz says. “When we confronted this person, they literally hung up the phone and we never heard from that person again.” The fraudster likely had three to four people writing code for them, he says, and worked at Jurkiewicz’s company for about five months before being caught.
“As a small company, it threatened our business, but you can imagine, there are some very large companies out there that employ hundreds of thousands of engineers, and this level of fraud could cost them millions, maybe even into the billions, if the scale is big enough and it goes on long enough,” Jurkiewicz adds.
Here’s another example, courtesy of CNBC: Pindrop Security, an Atlanta-based voice authentication startup, hired someone for a senior engineering role, but problems arose when the candidate’s facial expressions and word use appeared mismatched during video calls. And that’s not an uncommon occurrence.
Ben Sesser, CEO of BrightHire, an interview intelligence platform, shared an experience in which a client looked one way during a Zoom call, but the screen flickered and audio appeared mismatched with the person’s face. The candidate was asked to turn off their video and turn it back on—and when they did so, they looked completely different, due to deepfake technology.
Sesser explains that fake candidates will boast about their experiences working for tech companies like Amazon or Meta that are “too good to be true.”
“If you just think about what the approach is for these folks, it's very rote,” Sesser says. “It's very many applications, perfect résumés.”
Fraudsters often target entry-level and midlevel engineering, along with tech roles such as network administrators, Jurkiewicz says: “The real threat to employers is not just the fact that someone can be hired, but it's also the fact that during the hiring process, time is being wasted.”
How Can Authentic Candidates Succeed Against Fake Candidates?
If authentic candidates are using AI, they should openly disclose that during a job interview. That can give them an advantage, according to Jurkiewicz. In fact, it would be unrealistic for workers to not use AI in some capacity during the job-hunting process, he notes.
Jurkiewicz says candidates can also distinguish themselves from fake candidates by not only highlighting skills on their résumé, but explaining how they were developed and how the skills are strategic to their potential employer. Selling how you can use those skills in a specific organization could go beyond anything AI can do, Jurkiewicz suggests.
In addition, candidates can have potential employers call a reference at their corporate number rather than just a mobile phone number to prove that their experience is legitimate.
To counteract fake candidates that pose as competition, authentic candidates could show up to at least one interview in person, especially for critical or sensitive roles, Karnik advises: “Companies can and should continue hiring remotely, but for sensitive positions, it may be worth requiring at least one key interview to happen in person — or adding other verification steps to reduce risk.”
In addition, showing a level of personalization during the application process could help an authentic candidate stand out from a large pile that includes fake candidates. “Anything that a job seeker can do, whether it's in the application or in the interview process, to stand out, just to show a level of effort and personalization is going to be important,” Sesser says, “and a clear delineator between somebody who's putting some time and attention into the endeavor versus these folks who are applying for thousands of jobs.”
