Cybercrime is constantly using advanced techniques to attack users, their computers and phones, so much so that the latest criminal method hackers are using is Artificial Intelligence.
In addition to its implementation, it is possible to convert a hypothetically harmless application or file into malware capable of attacking slowly and silently. This threat is capable of bypassing antivirus systems and is activated at a specific time, when the victim executes a specific action. This malicious program is capable of learning to cause the greatest possible damage thanks to the data collected. In addition, it could be activated by geographic location, facial recognition, voice recognition, among other ways.
Cybercriminals will also store and analyse information about the victim’s behaviour in order to emulate it and gain access to different parts of their online activity. They will even create personalised attacks, as they will understand their interests, habits and social groups. Malware developers will always be looking for new vulnerabilities and today they can use Artificial Intelligence for their own purposes.
Ultimately, this intelligent malware will be one of the most widespread threats in the coming years. For some time now, AI has been increasingly present in our lives, starting with domestic environments with voice assistants and connected devices, and continuing with its deployment in organizational processes. For this reason, we must be prepared and, according to Panda Security, the three key cyberattacks that will be used the most are the following:
- Bots with the voices of famous people: Last year, the company Lyrebird created an algorithm that can imitate the voice of any person who speaks in the same way and uses the same verbal expressions. There is a possibility that hackers will take advantage of this license and improve the algorithm to create bots that imitate people more accurately. Thus, a bot of this type could call a database of people and obtain passwords, access codes to bank accounts and credit card numbers.
- Chatbots and emails: Many companies have adopted communication with their customers through chatbots, either on websites or through mobile applications, as well as by sending emails with specific content, so a large number of attacks could come from some vulnerability in these points.
- Mobile malware: Smartphones are adopting AI through the new chips that are being developed. As a result of this adoption, the device will be able to interpret certain repetitive actions of its owner in order to predict the demand for use and services. Shopping preferences, meals or restaurants visited generate quite useful Big Data for AI algorithms to offer products and services appropriate to the user’s needs. It is a double-edged sword if cybercriminals take control of the device.
The recommendations to protect ourselves from these dangerous AI-based malware are very similar to the general advice to protect ourselves from any type of cyber threat: it is vital to keep systems updated and the latest patches installed to minimize risks. However, the most important thing is common sense and making sure not to download files or install programs before checking them, as they may seem legitimate, but they may not be.
