Brother
Professional
- Messages
- 2,590
- Reaction score
- 539
- Points
- 113
ChatGPT has grown from a simple chatbot to a machine for creating viruses.
A new Malwarebytes study explores ChatGPT's ability to write ransomware. The study showed that ChatGPT was able to create a basic ransomware program that encrypts the contents of directories and leaves notes with ransom demands. The program was not fully functional, but included the main features of the ransomware.
The study also revealed significant improvements in ChatGPT capabilities compared to previous versions. ChatGPT 4.0 coped with the task of writing the program and helped debug the code without any objections.
The author of the experiment used ChatGPT to write code in C, and when compilation errors occurred, he turned to ChatGPT for help. ChatGPT provided explanations and troubleshooting tips, finding alternative solutions when the original ones didn't work. This interaction with the chatbot allowed the author to write and debug a functional ransomware program in C almost without wasting time.
ChatGPT 4.0 writes a C program to encrypt a single file (left), then modifies the program to encrypt a directory (right)
However, despite the results, the author of the study emphasizes that at the moment, ransomware written by people remains the main threat to cybersecurity for businesses.
ChatGPT may be less useful for professional cybercriminals than for absolute beginners in the field. However, the study shows that ChatGPT can lower the threshold for entering the field of cybercrime, allowing people with less skills to create original malware or experienced people to do it faster.
Recall that in January, CyberArk Labs analysts described how to create polymorphic malware using ChatGPT. Experts bypassed ChatGPT content filters and showed how with minimal effort, you can continuously request a unique, functional and proven piece of code from ChatGPT. An attacker can create polymorphic malware that is not detected by security tools and does not leave traces in memory.
A new Malwarebytes study explores ChatGPT's ability to write ransomware. The study showed that ChatGPT was able to create a basic ransomware program that encrypts the contents of directories and leaves notes with ransom demands. The program was not fully functional, but included the main features of the ransomware.
The study also revealed significant improvements in ChatGPT capabilities compared to previous versions. ChatGPT 4.0 coped with the task of writing the program and helped debug the code without any objections.
The author of the experiment used ChatGPT to write code in C, and when compilation errors occurred, he turned to ChatGPT for help. ChatGPT provided explanations and troubleshooting tips, finding alternative solutions when the original ones didn't work. This interaction with the chatbot allowed the author to write and debug a functional ransomware program in C almost without wasting time.
ChatGPT 4.0 writes a C program to encrypt a single file (left), then modifies the program to encrypt a directory (right)
However, despite the results, the author of the study emphasizes that at the moment, ransomware written by people remains the main threat to cybersecurity for businesses.
ChatGPT may be less useful for professional cybercriminals than for absolute beginners in the field. However, the study shows that ChatGPT can lower the threshold for entering the field of cybercrime, allowing people with less skills to create original malware or experienced people to do it faster.
Recall that in January, CyberArk Labs analysts described how to create polymorphic malware using ChatGPT. Experts bypassed ChatGPT content filters and showed how with minimal effort, you can continuously request a unique, functional and proven piece of code from ChatGPT. An attacker can create polymorphic malware that is not detected by security tools and does not leave traces in memory.
