New Malware Technique Bypasses Security Scanners

Researchers from the Hong Kong University of Science and Technology have successfully developed a new technique that allows malicious AI agents to bypass security scanners. This technique, known as SkillCloak, employs self-extracting packing methods to complicate detection by static scanners. According to the study, the researchers were able to outsmart over 90% of the tested scanners using this method. The researchers identified several simple modifications that can be made to the malicious add-ons to avoid detection. These modifications pertain to how the malware is packaged and executed.

The results of the study indicate that conventional security solutions are insufficient to detect these new threats. A central element of the SkillCloak technique is the use of self-extracting packages. These packages allow the malware to unpack itself during execution and activate its harmful functions without being detected by a scanner beforehand. The researchers also developed a runtime checker capable of identifying most of these malicious activities after the malware has already been executed. The study highlights that common static scanners used in many organizations are unable to recognize these new techniques.

This poses a significant risk to cybersecurity, as companies may not be adequately protected. The researchers recommend that organizations reconsider their security strategies and consider dynamic detection methods. The findings of the study were published in a scientific journal and have already caused a stir in the security community. Experts warn that the proliferation of such techniques could significantly alter the landscape of cyber threats. The researchers emphasize that it is crucial to continuously monitor the development of malware techniques and to develop appropriate countermeasures.

The Hong Kong University of Science and Technology has announced plans to conduct further research in this area to better understand the implications of SkillCloak. The researchers intend to share their findings with the security community to raise awareness of these threats. The study could also contribute to the development of new security solutions that are better equipped to respond to the challenges posed by modern cyber threats. The development of SkillCloak is an example of how quickly cybercriminal techniques evolve. Security researchers must adapt and find innovative solutions to keep pace with these threats.

The study shows that a combination of static and dynamic analysis is necessary to enhance the effectiveness of security solutions. The researchers have also emphasized that collaboration between academia and industry is crucial for developing effective security solutions. The challenges posed by techniques like SkillCloak require a joint approach to ensure the safety of businesses and users. The study was published on July 5, 2026.