Using generative AI to create cyber scams expected to explode in 2024

According to Kaspersky’s forecasts, the use of productive Artificial Intelligence (AI) platforms for cyber fraud is expected to explode in 2024. ChatGPT, Google Bard, Copilot, Midjourney and other services can be used by cybercriminals in more sophisticated attacks.

At an event held this Tuesday (21), experts from the digital security company announced that the use of prolific artificial intelligence tools for crime already exists. However, there is an expectation that there will be a large increase in its use by criminals due to its widespread use in 2023.

“Any cybercriminal with basic knowledge will be able to use increasingly accessible tools,” said Fabio Assolini, Director of Kaspersky’s Global Research and Analysis Team for Latin America.

ChatGPT is unlikely to be used as a murder weapon, but itself and its competitors can be used to create fake ads, emails and websitesFor example.

Experts argue that thanks to the improved systems, real content is almost indistinguishable from fake, as people “learn” the tools every time they use them.

Using AI to code malware

Fabio Assolini gave an example of another use of artificial intelligence in creating new fraud. Cybercriminals will use more targeted tools that are not as popular as ChatGPT. create and improve malware lines of code.

He noted that artificial intelligence for programming has advanced a lot this year and that it is now possible to create code that is valid against malware and phishing. There are even special AI codes and plugins to steal credentials, for example.

“We have known that cybercriminals have been using automation for a long time, this is nothing new. But we will start to see more closed vehicles that are limited to certain groups of attackers,” Assolini said.

Anderson Leite, a security analyst at Kaspersky in Latin America, added that all these tools will be important for what is called “polymorphy,” for example. Polymorphic malware can change its appearance to resemble other softwareAll this to avoid detection by security programs.

“For example, we have already seen a lot of use of AI for deepfake fraud and voice to assist in social engineering over the phone. But we still have a lot of room in the code to see what the impact will be,” Assolini said.

Top cyber crimes of 2024

The explosion in the use of productive AI in cybercrime is just one of Kaspersky’s predictions. Based on previous years’ trend reports and research on new types of fraud that have recently emerged, the company has created a list of eight other types of fraud that need attention in 2024.

Check it Other predictions from Kaspersky for next year’s cybercrime scenario:

• Increase in fraud schemes targeting the A2A system (Pix and others);
• Global adoption of Automated Transfer System (ATS) in mobile banking;
• worldwide spread of Brazilian banking trojans;
• Greater selectivity in choosing victims of ransomware attacks;
• Leveraging open source programs to carry out attacks against companies;
• Greater exploitation of one-day vulnerabilities at the expense of zero-day exploits;
• More attacks against misconfigured devices and services.