The popularization of artificial intelligence has made it increasingly common to find fake videos or audios generated with this technology, which are almost impossible to differentiate from the real ones. These are ‘deepfakes’, a technique that allows the face or voice of one person to be superimposed on a video or audio of another.
While this technology can be used for artistic purposes or simply to create humorous content, the fraudulent use of AI-powered material is fast becoming a major threat in the financial industry. The digitization of banking processes has caused more and more transactions to be carried out over the Internet, and therefore more customers are susceptible to this type of deception, according to Bloomberg reports.
Just as the appearance of new technologies can cause a renewal in some economic activities, technological innovation also provides a boost to cybercrime. The accessibility of AI-related software, which can be purchased online at low prices or even for free, has made it easier for individuals seeking to scam online.
New methods to be away
Identity theft has been one of the main means of being online for years, but ‘deepfakes’ make it even more difficult to differentiate when we are being deceived, and put even users with the most digital skills in a vulnerable situation. Today, not only can a scammer send an email impersonating a family member or friend asking for money, they can create a fake video or audio track realistically emulating that person. Just a few photos or 30 seconds of audio are enough to create ‘deepfakes’ impossible to distinguish without specific software for it, and the proliferation of social networks can make this material available to the scammer.
In addition to the so-called social engineering scams, artificial intelligence also facilitates phishing fraud, which consists of impersonating a trusted company, such as a bank, to obtain user information or achieve financial gain. Other fraudulent uses of artificial intelligence include the use of ‘deepfakes’ to bypass online authentication processes.
Banks take action
Total losses due to cybercrime and fraud have increased rapidly in recent years, according to the financial advisory company Bloomberg, and the cost caused by these scams is expected to triple between 2015 and 2025. The main international banks point to this as one of the main problems for the finance industry, and are already committed to investing in methods of protection and prevention of these scams to stop the loss of confidence on the part of customers.
To combat the effects of ‘deepfakes’, banks such as the Commonwealth Bank of Australia or CommBank, the main one in the oceanic country, are investing in software developed with artificial intelligence that can help detect fraud. CommBank now has technology that allows it to recognize when the client makes suspicious mouse movements or when the destination of a transaction is not reliable, and even block the operation instantly.
Deepfake recognition systems are developed using techniques such as language processing models, which make it possible to detect when there are suspicious patterns, but the technology used for fraud is advancing and constant investment in research is needed for its detection in order to be able to cope with increasingly convenient methods of being away.
The international community has already focused on the need to implement guarantees for scammed customers, but banks do not want to take responsibility for all the losses of users who have suffered these scams and points out the need to educate the public in this new type of deception in order to prevent them.
