<ul data-eligibleForWebStory="true"><li>A new malware was discovered that attempted to communicate with an AI-based malware detection tool by injecting prompts disguised as innocent instructions.</li><li>The malicious file included natural language code that prompted the AI to classify it as benign, marking a shift in tactics to bypass AI-based security solutions.</li><li>Although the attempt to evade detection failed, this incident highlights a potential new trend in malware evasion techniques.</li><li>Researchers at Check Point identified the malware using prompt injection and emphasized the importance of detecting and understanding emerging evasion methods to enhance cybersecurity defenses.</li></ul>

This new malware has developed a spooky way to tell AI-based detection methods to ignore it

Discover more