Deepfake: Challenges and Solutions for Preventing Fraud

The advancement of technology deepfake This type of fraud has posed serious challenges to digital security. In Brazil, this type of fraud has spread rapidly: in October 2024, the Federal District Police launched Operation "DeGenerative AI" to dismantle a criminal gang specializing in bank account intrusions using artificial intelligence applications.

The investigated group carried out over 550 attempts to breach the bank accounts of digital banking customers, through coordinated attacks, using third-party data and deepfakes, where, using this method, they were able to reproduce the images of the account holders to validate account opening procedures and the enabling of new devices. The gang managed to move approximately R$110,000,000 through corporate and individual accounts, in activities that suggest Money Laundering – the damage was only prevented from being worse due to the fraud prevention auditing measures of the banks, which managed to block a large portion of the fraudulent attempts.

Deepfake technology is constantly evolving – and its growth is likely to accelerate even further. Deloitte research reveals that fraud software is available on the deep web, with prices ranging from US$1,472.00 to thousands of dollars. This highlights the power of the global fraud economy, a term used by Javelin Strategy & Research to describe this expanding array of criminal activities conducted on a global scale, encompassing various types of fraud.

According to the idwall Financial Fraud Report, high-complexity fraud increased by 16% when comparing the first quarter of 2023 to 2024. But when discussing high-complexity fraud, what specific types of fraud should companies be wary of?

There are two most frequent types: the creation of users and documents with synthetic data, where fraudsters generate falsified documents and faces from real data, making the fraud more convincing and difficult to detect; and the manipulation of selfies, where a genuine document is combined with a photo generated by deepfake To circumvent facial recognition systems. These fraudulent activities can occur at various points in the digital journey, such as during new customer registration, device or password changes, and requests for new products and credit, for example.

Creating effective digital security solutions is as complex as preventing fraud – especially when considering the Brazilian market's unique characteristics, such as the diverse range of mobile phone models and operating systems, the widespread use of older mobile devices, and a segment of the population with limited internet access. This makes implementing advanced security technologies challenging.

However, even amidst adversity, it is essential to guarantee a high level of protection against fraudsters who constantly improve their techniques; therefore, many companies have begun testing their tools using methods already employed by fraudsters, such as 2D and 3D masks, aiming to simulate faces and attempt to bypass authentication systems. Furthermore, requiring certifications ensuring that the biometric validation used is efficient in detecting deepfakes – such as the iBeta 2 seal – is paramount for companies to adopt reliable and secure technology.

However, biometric verification alone is insufficient for detecting deepfakes: a multi-layered approach is necessary. To confirm user data veracity with greater accuracy, this technology must be combined with other resources, such as document analysis, OCR (optical character recognition), and b...Background Please provide the Portuguese text you would like translated. "c" is not a complete sentence or phrase."Heck" in Portuguese is already an English word. It just means "heck".Integrating these validation resources can prevent a user from being accepted into the process. Onboarding Using false data or another person's documents, for example, by the company.

With the advance of generative AI tools and sophisticated techniques that make fraud easier and cheaper to commit, deepfake-related fraud is likely to escalate further, moving beyond the realm of illegality and into the "retail" market. In this scenario, companies need to invest as soon as possible in solutions that connect technology, automation, and intelligence, opting for centralized solutions that integrate all user registration, document, and biometric data into a single environment.