A scintillating new report by online ID verification and cybersecurity company Sumsub suggests that 2023 has been the year that saw a massive increase in the leveraging of advanced, professional-grade fraud tactics.
According to the report, over 70% of crypto companies have observed new patterns and schemes adopted by bad actors, especially deep fakes, which pose a significant threat to these entities. With enhancements in artificial intelligence (AI) technologies, companies are starting to note a worrying uptick in the use of deep fakes by scammers to defraud their victims.
AI Advancements Lead to Hackers Using Deep Fakes To Bypass Security Measures of Crypto Platforms
What is even more concerning is the fact that compared to the previous year, there was a staggering 128% increase in deep fakes being used within the crypto industry in 2023. The report also shows that 55% of crypto service providers experienced a spike in fraud-related losses that have severely impacted both their financial and reputational standings.
Sumsub’s report pointed out a couple of instances where fraudsters used AI to exploit databases and launder stolen funds in crypto. In one case, a crypto company reported two AI-generated user profiles on their platform that were sophisticated enough to evade detection by the company’s cybersecurity protocols. The scenario highlighted the necessity of incorporating diligent manual ID verification protocols on crypto platforms.
In another case, a massive layering scheme affected the users of a popular financial platform in France, Italy, Portugal, and Spain where fraudsters attempted to launder funds through bank transfers to mules who then converted the money into crypto. The funds were then sent to a presumably popular cryptocurrency exchange in a deliberate attempt to obscure its origin.
The onboarding Processes of Crypto Platforms are Significantly Faster and More Efficient
However, the report was not all critical of cryptocurrency platforms. Sumsub highlighted a major shift taking place within the industry when it comes to users’ ID verification where platforms are swiftly moving from document-based to non-document-based verification procedures. This has heralded in faster verification times that have gone from an average of 38 seconds to sometimes as quick as under five seconds.
Verification times have nearly halved from 2022 levels and this is despite the proliferation of AI-generated profiles this year. European platforms are considered to have the fastest onboarding times, taking less than 30 seconds on average to confirm the eligibility of a new user. Verification times have also improved in Africa, Asia, and Latin American regions.
Sumsub also observed a significant decrease in average verification times among its crypto clients, especially in Brazil, Mexico, and Germany, where there was a massive improvement in customer onboarding procedures.
AI-powered Verification Tools Necessary to Detect Fraud Tactics Tailored For Modern Era
However, there is another challenge faced by verification providers. Despite becoming highly efficient, they are troubled by Proof-of-Address (PoA). Since document types can vary from country to country, it has led to a drop-off in new client registrations as users are struggling to submit the required documents.
The situation reportedly arose because legacy know-your-customer (KYC) providers are finding it difficult to identify advanced fraud techniques, such as AI-generated profiles, further emphasizing the need for robust and technologically advanced solutions.
Crypto companies are finding a way around this problem by seeking the help of companies that employ advanced technical infrastructure that is well capable and experienced in handling extensive amounts of documents.
Analysts at Sumsum conclude their report by suggesting that leveraging AI-powered modern anti-fraud solutions is necessary to combat new-age fraud methods. These advanced programs are capable of detecting deep fake products, searching for varied document templates by conducting a similarity analysis, and automatically rejecting applications that are flagged invalid.