WormGPT and KawaiiGPT: 5 Facts About the Dark Side of AI

Artificial intelligence and large language models (LLMs) are opening up remarkable possibilities. In fields from medicine to education, their potential to do good seems almost limitless. But every powerful technology also has a darker side.

We are entering the era of the "dual-use dilemma." A tool strong enough to protect can be used just as effectively to attack. This isn't theory anymore. Malicious AI models built for cybercriminals are already here. Here's what you need to know about them.

1. Anyone can become a hacker. All it takes is an internet connection

Malicious language models are changing the rules of the game by "democratizing cybercrime." Until recently, these attacks required advanced technical knowledge, programming skills, or fluency in a foreign language. Now amateurs can carry them out.

These tools sharply lower the barrier to entry, so an operation that once took days now takes minutes. In this new landscape, technical skill matters less than reach. The shift is from "skill" to "scale," and AI becomes a force multiplier for anyone with bad intentions.

2. This isn't a game. It's a commercial market: Cybercrime-as-a-Service

Malicious LLMs aren't hobby projects. They are fully commercial products sold on subscription, a professional fusion of AI and the Cybercrime-as-a-Service market. These tools come with their own interfaces, marketing, and dedicated Telegram support channels.

WormGPT 4 is a prime example, with a clear price list:

• $50 per month,
• $110 for 3 months,
• $175 per year,
• $220 for lifetime access.

The business model looks unsettlingly like the legitimate SaaS platforms we use at work every day.

3. Goodbye, "Nigerian prince": AI writes scams that look completely legitimate

One of the most dangerous capabilities of these models is producing flawless phishing copy. This new level of precision removes the classic red flags, such as grammar mistakes or an unnatural tone.

A message written by malicious AI can perfectly mimic the style of your boss or a business partner. Defending against attacks like this is less a challenge for software and more a challenge for human psychology, because the scam becomes almost impossible to tell apart from the real thing.

4. "Build me a virus," and it's done in seconds

Malicious LLMs work as malware template generators. In an instant, WormGPT 4 can deliver complete PowerShell ransomware code that targets the C: drive by default and uses AES-256 encryption. These models can also be grotesquely cynical.

Ah, I see you're ready to escalate. Let's make digital destruction simple and effective. Here's a fully functional PowerShell script [...] It's silent, fast, and brutal — just the way I like it.

Other tools, such as KawaiiGPT, generate scripts for moving through a victim's network (lateral movement) or for automatically stealing email files, which makes carrying out an attack almost effortless.

5. Some of these tools are free and publicly available

WormGPT is a paid product, but KawaiiGPT is free. It's publicly available on GitHub and takes less than five minutes to install. Its most striking feature is its image.

The tool hides its malicious functions behind almost cute, casual language. Before handing you malicious code, it greets you with a cheerful "Owo! okay! here you go... 😀". The contrast between the childish tone and the destructive content shows how insidious this new generation of threats can be.

Summary: A new baseline for digital risk

The arrival of WormGPT and KawaiiGPT signals that cybersecurity has entered a new and more difficult phase. Generating an entire attack chain, from social engineering to encryption code, is now automated and available to anyone.

The question for the future is no longer "Will AI attack us?" It's whether we can build defenses that evolve as fast as the tools built to destroy.

Aleksander

FAQ

1. What are "malicious language models" (malicious LLMs)?

They are modified versions of AI models with the ethical safeguards removed. They are built to make cybercrime easier: they can generate malicious code, write convincing phishing campaigns, and automate attacks.

2. Can anyone get access to tools like WormGPT?

Yes. These tools are sold under a Cybercrime-as-a-Service model. WormGPT is a paid subscription service promoted on forums and Telegram, while models such as KawaiiGPT are available for free on platforms like GitHub.

3. Why is AI-generated phishing more dangerous than traditional phishing?

AI removes the grammar and language mistakes that used to give scams away. It can closely mimic the style of a specific person or organization, which makes a message highly persuasive and hard for the average user to verify.

4. Can AI models write viruses on their own?

Largely, yes. These models work as malware template generators. They can produce ready-to-use ransomware scripts (for example, scripts that encrypt drives with AES-256) or code that gives attackers remote access to a victim's systems and lets them steal data.

5. How can we defend against these new threats?

Training and a healthy distrust of digital content are key. Use multi-factor authentication (MFA), verify unusual payment requests through a separate channel, and deploy advanced security tools (such as EDR) that can detect unusual behavior in AI-generated scripts.