Generative AI technology is rapidly growing and advancing, driven by continuous research and development efforts. However, these generative AI technologies are also actively exploited by threat actors for their illicit activities.
Recently, cybersecurity researchers from SlashNext reported that threat actors are actively using ChatGPT and a black-hat alternative to GPT models, WormGPT, to launch business email compromise (BEC) attacks.
WormGPT is revolutionizing BEC attacks. ChatGPT and other advanced AI tech enable threat actors to automate convincing personalized fake emails, expanding the scope of BEC attacks and boosting the attack’s success. These advanced AI technologies are indirectly aiding the threat actors to craft sophisticated phishing emails, surpassing language barriers and enhancing their attack effectiveness.
In these types of events, threat actors exploit the interfaces of the ChatGPT or similar tools with specialized prompts to manipulate and compromise AI. This is clearly highlighting the urgent implementation of robust AI security measures. Moreover, to reach the next level, cybercriminals also develop user-friendly custom AI modules similar to ChatGPT for illicit use, amplifying the complexity of cybersecurity in an AI-driven world.
WormGPT is mainly designed for malicious activities since it’s a malicious alternative to GPT models. Apart from this, WormGPT offers several powerful features, such as unlimited character support, chat memory retention, and code formatting. WormGPT trained on undisclosed malware-related datasets, keeping its training sources confidential as per the author’s decision.
Researchers thoroughly evaluated the risks of WormGPT by testing its ability to generate a convincing threatful email targeting an unsuspecting account manager for payment of a fraudulent invoice. The disturbing outcome revealed the exceptional persuasive and tactful email generation capabilities of WormGPT, demonstrating its dangerous capabilities to generate advanced phishing emails and launch BEC attacks.
WormGPT is an unrestricted variant of ChatGPT since it lacks ethical boundaries or limitations, unlike ChatGPT. WormGPT highlights the significant risk of generative AI.
The advantages of generative AI for BEC attacks include exceptional grammar and a lowered entry threshold.
The article ends with some recommendations offered by the security analysts:
- BEC-Specific Training
- Enhanced Email Verification Measures
- Make sure to test your security efficacy in observability mode
- Always use a robust security solution.
