Threat factors that undertake identity fraud used in 2023 ten times more than in 2022, according to the SumSub’s digital identity verification solutions.
In his third annual Identity fraud reportPublished on November 28, 2023, Sumsub discovered that artificial intelligence-driven techniques were among the top five tools that were used to perform fraud online in 2023.
The increase in deep fakes was good for most AI-driven fraud techniques recorded by the company.
Spend against Infosecurity During a Sumsub AI regulation event held in London on November 23, Ilya Brovin, the chief growth employee of the company, warned about the growing refinement of identity fraud.
“Yes, the AI and DeepFake tools become easier to use and more accessible to everyone. But also, many fraudsters who use AI-driven tools are not the random individuals that we sometimes saw in identity fraud, they are part of organized groups,” he added.
Where has fraud the most increased?
Although globally, the increase in the use of deepfake was substantial in North America, where in 2023 it increased by 1740% compared to 2022, and in Asia-Pacific (APAC), which saw an increase of 1530% in the same period.
The use of deep food for identity fraud has risen by 450% in the Middle East and Africa (MEA) and 410% in Latin -America.
The Cryptocurrency sector is most focused on AI-driven fraud techniques that represent 88% of all Depfake cases that are detected by Sumsub in 2023, followed by Fintech (8%).
In a public explanation, Pavel Goldman-Kalaydin, head of Sumsub of AI/ML, noted: “The rise of artificial intelligence reforms how fraud is committed and prevented. AI serves as a powerful tool for both anti-fraud solutions and that of identity fraud.”
“Deepfakes pays the way for identity theft, scams and wrong information campaigns on an unprecedented scale,” he added.
Read more: Cyber criminals hesitate about the use of generative AI
Bangladesh and Pakistan, top goals of identity fraud
In general, the worldwide rate of identity fraud almost in 2023 doubled from 2021, when Sumsub started its annual fraud report.
Bangladesh is the country where identity fraud is most common in the telemetry of Sumsub, accounting for 5.44%of detected fraud in the 2023 database, followed by Pakistan (4.59%) and Latvia (4.05%).
Online media is the most targeted industry in general, with 4.27%of the detected fraud in 2023, followed by professional services (3.14%) and healthcare (2.41%).
What are the top fraud techniques in 2023?
The other top fraud techniques observed in 2023 included:
- Reporting mulation networks
- Fake -ids
- Account takeovers
- Forced verification
Money Muling refers to fraud campaigns in which apparently innocent persons, known as money mules, are recruited to transfer illegally obtained funds, causing their origin to be disguised.
Forced verification, which happens when individuals are manipulated to go through Know-Your-Customer (KYC) processes for fraudsters, 305% grew above 2022-2023.
Two new emerging trends
Moreover, Sumsub warned of two emerging trends that it had never seen before.
Firstly, the researchers of identity fraud discovered that documents of developed economies are increasingly forged by fraudsters.
“Because documents from developed economies are often seen as more credible, they become new goals for fraudsters. That is why it is necessary for organizations and individuals to recognize the need for improved document verification, even from apparently reliable sources,” the report is.
The second trend is that non-regulated entities are more often aimed.
“In an era in which legal frameworks play a crucial role in shaping security practices, non-regulated entities are becoming increasingly susceptible to the developing tactics of fraudsters.
Finally, Sumsub encouraged anti-fraud companies to make their technologies more accessible by exchanging knowledge and experience to “enable the entire industry to stay ahead of increasingly advanced fraud tactics.”
The Sumsub Identity fraud report is based on aggregated and anonymized verification statistics on millions of users from 28 industries worldwide, with more than 2 million fraud attempts studied.