By Dr. Tim Sandle
SCIENCE EDITOR
DIGITAL JOURNAL
April 16, 2026
Activists hold eye masks during a protest against deepfake porn in Seoul - Copyright AFP Anthony WALLACE
The ‘deepfake phenomenon’ began circa 2019 and slowly something that hit most members of the public as an amusing tool has become a dominant form of cybercrime.
In terms of digital and non-digital boundaries, deepfake technology has seemingly created a ‘new reality’ where criminals are merrily cashing in on an unsuspecting public on a significant scale. This is the trend from a new study by cybersecurity company Surfshark.
Deepfakes rely on a type of neural network called an autoencoder. These consist of an encoder, which reduces an image to a lower dimensional latent space, and a decoder, which reconstructs the image from the latent representation.
The study also breaks down the full financial toll, revealing the hardest-hit countries and the most effective scam tactics fuelling this digital crime wave. Overall, deepfake fraud has already cost $2.19 billion globally. This is a figure that is set to grow.
Faked celebrity endorsements
In terms of the most powerful forms of deepfake, investment endorsements by celebrities and government officials top the list (52% of all reported deepfake-related fraud losses), costing $1.13 billion.
This is followed by corporate attacks — such as the impersonation of CEOs to request unauthorized transactions — at 25%. Other significant contributors include financial crimes where victims’ identities are stolen and scammers use deepfake technology to secure bank loans or drain accounts (9%), followed by deepfaked romance scams (7%), family member impersonation (6%), and various other forms of deepfake-related fraud (2%).
U.S. and Europe lead the victims
Geographically, the U.S. is the most targeted country globally for deepfake-related scams, suffering $712 million in losses. Of these, 43% occurred in the corporate sector — involving scams in which deepfakes were used to trick organizations into sending money or, in some cases, to place fake candidates in remote jobs. Another 31% of losses resulted from deepfaked investment opportunities.
Europe’s top deepfake fraud victims are the UK ($149 million), Sweden ($63 million), and Spain ($56 million), each ranking in the global top 10. In these three countries, 90% of losses were caused by scams that used deepfakes of famous people to endorse various investment opportunities. The remaining losses, particularly in the UK, were due to scams targeting corporations and incidents involving deepfake romance schemes.
With other top-ranking countries Malaysia, ranking second globally with $502 million in losses, was almost entirely targeted by deepfaked investment schemes (99.7%).
Meanwhile, Hong Kong, in third place with $229 million in losses, has become the global hub for deepfake romance fraud ($105 million).
However, Indonesia stands out as a major outlier. Ranking fifth with $139 million in losses, nearly all of its damages came from criminals using deepfake technology to bypass bank security measures to secure fraudulent loans.
Activists hold eye masks during a protest against deepfake porn in Seoul - Copyright AFP Anthony WALLACE
The ‘deepfake phenomenon’ began circa 2019 and slowly something that hit most members of the public as an amusing tool has become a dominant form of cybercrime.
In terms of digital and non-digital boundaries, deepfake technology has seemingly created a ‘new reality’ where criminals are merrily cashing in on an unsuspecting public on a significant scale. This is the trend from a new study by cybersecurity company Surfshark.
Deepfakes rely on a type of neural network called an autoencoder. These consist of an encoder, which reduces an image to a lower dimensional latent space, and a decoder, which reconstructs the image from the latent representation.
The study also breaks down the full financial toll, revealing the hardest-hit countries and the most effective scam tactics fuelling this digital crime wave. Overall, deepfake fraud has already cost $2.19 billion globally. This is a figure that is set to grow.
Faked celebrity endorsements
In terms of the most powerful forms of deepfake, investment endorsements by celebrities and government officials top the list (52% of all reported deepfake-related fraud losses), costing $1.13 billion.
This is followed by corporate attacks — such as the impersonation of CEOs to request unauthorized transactions — at 25%. Other significant contributors include financial crimes where victims’ identities are stolen and scammers use deepfake technology to secure bank loans or drain accounts (9%), followed by deepfaked romance scams (7%), family member impersonation (6%), and various other forms of deepfake-related fraud (2%).
U.S. and Europe lead the victims
Geographically, the U.S. is the most targeted country globally for deepfake-related scams, suffering $712 million in losses. Of these, 43% occurred in the corporate sector — involving scams in which deepfakes were used to trick organizations into sending money or, in some cases, to place fake candidates in remote jobs. Another 31% of losses resulted from deepfaked investment opportunities.
Europe’s top deepfake fraud victims are the UK ($149 million), Sweden ($63 million), and Spain ($56 million), each ranking in the global top 10. In these three countries, 90% of losses were caused by scams that used deepfakes of famous people to endorse various investment opportunities. The remaining losses, particularly in the UK, were due to scams targeting corporations and incidents involving deepfake romance schemes.
With other top-ranking countries Malaysia, ranking second globally with $502 million in losses, was almost entirely targeted by deepfaked investment schemes (99.7%).
Meanwhile, Hong Kong, in third place with $229 million in losses, has become the global hub for deepfake romance fraud ($105 million).
However, Indonesia stands out as a major outlier. Ranking fifth with $139 million in losses, nearly all of its damages came from criminals using deepfake technology to bypass bank security measures to secure fraudulent loans.
No comments:
Post a Comment