4 Dark LLMs: Black Hat Hackers are Also Riding the GenAI Wave The global market of GenAI is projected to reach USD 67.18 billion in 2024 to USD 967.65 billion by 2032

We are always reading about Generative Artificial Intelligence (GenAI) and how it is already revolutionizing every industry and the workplace. The global market of GenAI is projected to reach USD 67.18 billion in 2024 to USD 967.65 billion by 2032 from USD 43.87 billion in 2023 with a CAGR of 39.6 per cent during the forecast period (2024-2032).

No wonder, with you, black hat hackers are also riding the GenAI wave. Some have been observed hacking legitimate large language model-based (LLM) tools, while others have crafted their own malicious LLM model by leveraging the same principles as benign LLMs but with the intent to write malicious code, scam pages, and fraudulent messages, phishing pages, creating undetectable malware, and other hacking tools for hackers.

Here are four malicious LLMs which are proving to do more harm than good:

WormGPT

It is an open-source LLM created by EleutherAI in 2021 for malicious actors. It can only be accessed behind a paywall on a dark website—which allows the user to bypass some of ChatGPT's guardrails by injecting specific instructions in the prompt. This method is called 'LLM jailbreaking.' "Biggest enemy of the well-known ChatGPT that lets you do all sorts of illegal stuff," said EleutherAI.

WormGPT generates harmful and illegal content without any moderation, creates phishing emails, malware, and other malicious scripts, and spreads misinformation, hate speech, or other damaging narratives.

FraudGPT

FraudGPT or FraudBot was launched on the dark web and Telegram channels on July 22, 2023. According to Cybersixgill FraudGPT provides "exclusive tools, features, and capabilities" and "no boundaries."

It writes malicious code, fraudulent messages, scam pages, creating undetectable malware, phishing pages, and other hacking tools. Additionally, it finds leaks and vulnerabilities and monitors relevant groups, sites, and markets.

WolfGPT

Little is known about the WolfGPT, however, several hackers promoted the sale of this dark LLM. It was presented as an alternative to ChatGPT with malicious intent.

This model was built using Python and allegedly offers complete confidentiality, enabling powerful cryptographic malware creation and advanced phishing attacks.

XXXGPT

According to various sources, XXXGPT is designed to deploy botnets, remote access Trojans (RATs), and several types of malware tools such as ATM malware kits, cryptostealers, and infostealers for malicious actors.

Additionally, the developers of the XXXGPT claim that its tools are sophisticated and advanced, equipped with several revolutionary features and services.

In order to protect yourself from such malicious agents, do not click on random links provided in the emails and messages. Avoid sharing your sensitive credentials on social media such as your address, mail address, phone number, banking information, and Adhar information. Before clicking on any website, check its URL (actual URL: https, fake: http). Additionally, do not click on random ads. Also, keep your system updated.