28.07.2026 - Deepfake interviews featuring public figures are often used to promote fraudulent investment opportunities. They can also be used for bullying or blackmail by depicting someone in a compromising situation. In this instalment of its four-part summer series, the NCSC looks at face swapping. With the help of AI, it is now possible to replace faces in seconds with little effort and deceptively realistic results.
The previous instalments of this summer series discussed how fake images and text can mislead people and be used to commit fraud. One of the earliest and best-known examples of deepfakes is the use of fake interviews with celebrities to promote fraudulent investment offers. Images can also be used for bullying or blackmail, particularly when they depict a person in a compromising situation. Scams of this kind can be extremely distressing for the victim, even when the images are fake.
AI increases pressure on victims of fake sextortion
The NCSC regularly receives reports of fake sextortion cases, in which victims are threatened with the publication of personal information, images or videos unless they pay. In the vast majority of cases, these claims are baseless. Scammers send such emails indiscriminately, hoping that one of the recipients will pay the ransom. As a rule, however, they leave victims alone if they do not respond. The situation becomes more serious when scammers use AI to create fake images or videos of the victim. The NCSC has already seen a small number of such cases in which the fabricated material was used to make the threat more credible and increase pressure on the victim. Cases of this kind are likely to become more common as conventional fake sextortion scams become less effective. However, scammers can only create such material if they can find suitable images of the victim online. It is therefore advisable to limit the number of photos of yourself that are publicly accessible.
AI can be used for face or body swapping
In fake sextortion emails, cybercriminals often use AI to replace a person's face ("face swapping"), or even their entire body ("body swapping"), into another image or video. The results can be remarkably convincing. To spot these manipulations, pay close attention to subtle inconsistencies at the points where the images have been joined. The lighting and shadows on the inserted head often do not quite match those on the rest of the body. Other signs to look out for include slight blurring and differences in texture or image quality, such as a sharp face on a slightly pixelated body. Unnatural transitions around the neck or hairline can also reveal the manipulation.
One of the earliest and best-known uses of deepfakes has been to create fake interviews with celebrities and other public figures to promote fraudulent investments. Scammers were already using AI for this purpose three years ago, manipulating images of well-known people to make their offers seem credible and persuade victims to invest. In Switzerland, fake interviews featuring federal councillors and other well-known Swiss public figures are particularly common and are used to give dubious offers an air of legitimacy.
AI is helping existing scams evolve
Romance scammers have recently begun using AI-generated images. Previously, they relied on photos of uninvolved people copied from the internet, which limited the situations they could plausibly depict. Now, however, they can use AI to create entirely fictitious people. By generating images and videos, scammers can not only invent an entire identity, but also portray the person in situations that would previously have been impossible to fabricate. In a recent case, for example, scammers used AI-generated images to “prove” that their fake persona had been admitted to hospital. The NCSC covered the case in a previous weekly review.
AI-generated images are now also being used in conventional phishing attacks designed to take over accounts. In these cases, victims receive a message that appears to be from someone they know, asking them to vote for their child in a competition. Anyone who follows the link and unsuspectingly enters a verification code or scans the QR code on the scam website immediately loses access to their WhatsApp account. The voting page, including the photographs of the children, is entirely AI-generated. The NCSC reported on this in a weekly review.
Conclusion
Images can now be easily manipulated, so you should be cautious about what you see. Modern technology can be used to create entirely fictitious people or alter existing images to change body shape, height, and other things. Even a single photograph, for example from social media, can be used to insert a person or face into another image or replace someone already in it. This makes the standard warning signs of a scam even more important. Always consider whether the situation seems plausible. Scammers generally want one of two things: your login details or your money. Be extremely cautious whenever you are asked to hand either of these over.
Further information
- The evolving "romance scam" – Using AI to create the perfect deception
- Fake competition: Hackers hijack WhatsApp accounts with a voting scam
- Use of artificial intelligence in fraud attempts
- "Urgent money transfer" – CEO fraud remains a persistent threat to SMEs
- European Cyber Security Month (ECSM): Tips for senior users
- Report on the influence of artificial intelligence in cybersecurity
Current statistics
Last week's reports by category:
Last modification 28.07.2026
