Deepfake job applicants: how do you spot a fake candidate when hiring remotely?

A North Korean network got into more than 70 US companies with false identities. What that means if you hire remotely, and how to check who you are really hiring.

By Nikolay Broeks, Automation & Growth Engineer

More than 70 US companies hired IT workers who were not who they claimed to be, the US Department of Justice announced on 13 October 2026. The indictment points to a North Korean network. You rarely catch such a fake candidate on video alone, but you can catch the mismatch between interview, ID, address and bank account.

What is a deepfake job applicant?

A deepfake applicant applies under an identity that is not their own and uses edited or generated images or voice to make the interview match that identity. It is the digital version of sending someone else to the job interview.

The US case shows how far that can go. According to the indictment, North Korean IT workers applied for remote jobs from 2019 to around July 2025 using stolen or fabricated identity documents, and used VPNs to look as if they were working from inside the United States. Facilitators in the US received the laptops the companies shipped, installed remote access software and passed the salaries on. In late September the FBI carried out searches at more than 40 homes in Texas that were operating, or suspected of operating, as laptop farms, and seized around 187 laptops and other devices. Fourteen people have been charged. These are allegations: no one has been convicted.

The announcement does not say deepfakes were used in this particular case. Earlier warnings do describe tampering with image and voice. In June 2022 the FBI reported voice spoofing in interviews for remote IT and software roles, and lip movements on camera that did not fully match the audio. Germany's domestic intelligence service, the Bundesamt für Verfassungsschutz, wrote in 2024 that some application photos are generated with AI.

On 31 July 2026 the Netherlands joined a warning issued by eleven countries, including the United States, Germany and France. It says these workers are increasingly using AI to hide who they are.

How do you spot a fake candidate in a video interview?

Do not look only at the picture. The signals listed by the eleven countries and by the German service are mostly about things that do not fit together.

Face and voice. Video that looks edited or artificially generated, a face that does not match the photo on the ID, or lips that do not keep up with what you hear.

Who is on the other side. A candidate who refuses a video call, or an account where you seem to be talking to a different person depending on the time of day, because these workers operate in teams. The German service advises watching for eye movements and long pauses that suggest someone is reading out answers.

The CV and the profiles. A degree from one country and work experience only in a completely different one, social media profiles that do not match the CV, or several profiles with the same name and different photos.

Money and equipment. A rate below the market, a request to be paid in cryptocurrency or into someone else's account, and a delivery address for the laptop that keeps changing.

One signal on its own means little. A shaky connection or a sharp rate is true of plenty of genuine candidates. The international warning says so itself: only when several characteristics come together is there reason to look further.

How do you verify a candidate's identity when you hire remotely?

If you employ people in the Netherlands, you already have a fixed moment for this. Before someone starts, you establish their identity from the original document and keep a copy until five years after the year in which the employee leaves. That step stays with you as the employer, just like the VOG certificate of conduct, even when an agency does the recruiting. How we split the steps from intake to hire is set out in how we work.

When the whole process runs remotely, that moment easily shrinks to a scan sent by email. That is where the gap is. Three habits make it smaller.

See the person once, for real. The German service advises holding interviews in person or by video, and the international warning asks job platforms to require in-person interviews. Ask the candidate to show the original ID before the first working day, ideally at the office.

Check at the source. Have degrees confirmed by the institution that issued them, using contact details you looked up yourself rather than the ones on the CV. That advice comes from the German service. Approach previous employers the same way, as when you check references.

Ship the laptop to the right address. Only deliver equipment to the address on the ID, and make sure a new starter cannot install remote access software themselves. In October 2026 the FBI advised continuing to verify identity after the start date, not just during hiring.

If you hire for many remote roles at once, set the process up properly once. Even if you outsource recruitment through RPO, agree in advance who checks identity, and at which moment.

What do employers ask us about fake applicants?

Is this happening in the Netherlands too?

The Netherlands took part in the international warning of 31 July 2026, and the German service wrote back in 2024 that German companies had also worked with these IT workers. We do not know of a Dutch figure: we have not found any publication that counts the cases. The risk sits mainly in roles that are done entirely remotely.

Can I ask an applicant for a copy of their ID?

Not during the application process in the Netherlands. According to the Dutch Data Protection Authority, you may not ask for or make a copy, scan or photo of an applicant's ID, because it gives you data you do not need at that stage, such as the photo and nationality. You may ask them to show the original. More on how we handle candidate data is in data and privacy.

Does this only apply to IT roles?

Mostly, but the employers themselves come from many sectors. The US case hit companies in health care, communications, cyber support, retail and manufacturing. The 2026 warning names websites, apps, software and blockchain as the work these people look for. What the roles have in common is that they are done entirely remotely.

Should I ask my recruitment agency about this?

Yes, if the agency recruits for remote roles. Ask who checks candidates' identity, at what moment and with which document, and whether anyone has ever seen the candidate in person. Other questions worth asking are covered in choosing a recruitment agency.

SocialFind recruits for employers who need people on a structural basis, and wrote this article from the US Department of Justice announcement, the international warning of 31 July 2026, the warnings from the FBI and the Bundesamt für Verfassungsschutz, and the rules of the Dutch government and the Dutch Data Protection Authority, checked on 6 November 2026. If you would like articles like this in your inbox, sign up for our newsletter.

Our newsletter

What we run into in our work for employers, gathered in one email: new articles, labour market figures and what follows from them in practice.

About one email a week. You can unsubscribe with one click at the bottom of every email. What we do with your details is set out in our privacy statement.

We use cookies

This website uses cookies to work properly and to measure how the site is used. Statistics and marketing are only used with your consent. You can change your choice at any time via "Cookie settings" in the footer.