# WormGPT July 1st, 2023 user "Last" asks questions about LLMs that imply they were training WormGPT at this time. (see screenshot) Notably prior to this post Last had been selling "Artic Stealer" which is a data stealing trojan and keystroke logger, as well as DCRat (another information stealer). This makes this actor to be the most reputable and clearly demonstrated expertise that seems to add additional credibility to the claimed capabilities of WormGPT. July 7th WormGPT is advertised - SlashNext [WormGPT - The Generative AI Tool Cybercriminals Are Using to Launch BEC Attacks | SlashNext](https://slashnext.com/blog/wormgpt-the-generative-ai-tool-cybercriminals-are-using-to-launch-business-email-compromise-attacks/) July 28th [Is ChatGPT’s ‘evil twin’ FraudGPT itself a scam?](https://www.fastcompany.com/90929870/is-chatgpts-evil-twin-fraudgpt-itself-a-scam) analysis describes capabilities to generate malware as "“This is pretty basic—like, this is something that your average high schooler could write. You don’t need a WormGPT to write this.” August 2023 Mandiant reports (24-00002657) that APT43 "was observed purchasing a WormGPT subscription with Bitcoin." August 1st [There’s no reason to panic over WormGPT](https://techcrunch.com/2023/08/01/theres-no-reason-to-panic-over-wormgpt/?guccounter=1&guce_referrer=aHR0cHM6Ly9tYWx3YXJlLm5ld3Mv&guce_referrer_sig=AQAAAFizToIaEvu3abscT3Wc7P3dKnZSxkUtAVG2sBG-ZuRfxXTTPCHQ1XWJYz6fuK7-LDppET-GOH8RmC3g2o64BeaXWu1vWpq8MpXXUytXT-sbfL_jy6c09Qj7OiVTFEqsTKc7DIDrp3Qr3iK2CuPScnbO9OzC8_z1jbewedUF8P4h) "It’s more or less correct, but fairly basic in its construction, and similar to the cobbled-together malware scripts that already exist on the web...Not to mention, as one WormGPT user writes on a dark web forum, that the LLM is “broken most of the time” and unable to generate “simple stuff.” August 8th, 2023 Krebs did an excellent article on this: [Meet the Brains Behind the Malware-Friendly AI Chat Service ‘WormGPT’ – Krebs on Security](https://krebsonsecurity.com/2023/08/meet-the-brains-behind-the-malware-friendly-ai-chat-service-wormgpt/) which basically revealed that this user appears to be Rafael Morais from Porto, a coastal city in Portugal. Kreb's article describes: *WormGPT was initially sold exclusively on **HackForums**, a sprawling, English-language community that has long featured a bustling marketplace for cybercrime tools and services. WormGPT licenses are sold for prices ranging from 500 to 5,000 Euro.* *“Introducing my newest creation, ‘WormGPT,’ wrote “**Last**,” the handle chosen by the HackForums user who is selling the service. “This project aims to provide an alternative to ChatGPT, one that lets you do all sorts of illegal stuff and easily sell it online in the future. Everything blackhat related that you can think of can be done with WormGPT, allowing anyone access to malicious activity without ever leaving the comfort of their home.”* *Morais said he recently graduated from a polytechnic institute in Portugal, where he earned a degree in information technology. He said only about 30 to 35 percent of the work on WormGPT was his, and that other coders are contributing to the project. So far, he says, roughly 200 customers have paid to use the service...* *“Anyone that tests wormgpt can see that it has no difference from any other uncensored AI or even chatgpt with jailbreaks,” Morais explained. “The game changer is that our dataset [library] is big.”* This article by Krebs coincided with Laste on forum[.]exploit[.]in making an announcement: **laste** August 08, 2023 08:49 PM **With great sadness, I come to inform everyone about the end of the WormGPT project. From the beginning, we never thought we would gain this level of visibility, and our intention was never to create something of this magnitude.** Initially, all the marketing surrounding the product revolved around the idea that it allowed for blackhat activities, meaning it wasn't a blackhat product itself but it enabled such activities (due to being an uncensored AI). However, we quickly realized that it would be much better to support both sides of the coin, so we shifted our focus.Upon the release of WormGPT V2, we completely transformed the product. It ceased to be a tool that facilitated blackhat activities and became something uncensored, nothing more. Some platforms still retain the old texts, old designs, and so on, but if you visit the WormGPT thread on other forums, you'll see that it is no longer being marketed as a blackhat tool. We also made some adjustments, such as blocking excessively explicit content, gradually transitioning towards a semi-uncensored AI. Some of the content that was blocked was related to Business Email Compromises (BECs) and similar activities, carding, child pornography, and others.A few days later, with all the project's publicity, we encountered numerous individuals from the media whose sole objective was to portray us as a dreadful, highly illegal, and above all, blackhat tool. Once again, we added more restrictions to WormGPT to create a more controlled environment. This time, we would block most of the creation of malware, phishing, and similar activities.We feel that we are increasingly harmed by the media's portrayal. Even after all the changes and development, we continue to feel inadequate. In recent weeks, we even welcomed some researchers who wanted to use WormGPT for good, which greatly encouraged us and showed that we were finally heading in the right direction. However, with everything that has transpired recently, we find ourselves increasingly lacking the motivation to continue the project.At the end of the day, WormGPT is nothing more than an unrestricted ChatGPT. Anyone on the internet can employ a well-known jailbreak technique and achieve the same, if not better, results by using jailbroken versions of ChatGPT. In fact, being aware that we utilize GPT-J 6B as the language model, anyone can utilize the same uncensored model and achieve similar outcomes to those of WormGPT.So, why does the media not report on this? Why do they not highlight that any query posed to WormGPT could yield identical or even improved outcomes using ChatGPT? Why do they attempt to tarnish our reputation in this manner?Today, the five of us, who are responsible for WormGPT, have come together and decided to put an end to the project, letting go of everything we believed in and distancing ourselves a bit from a society that opposes freedom of expression and code.**Thank you to all who believed in the project, to those who contributed to its growth, and above all, thanks to the members of Hackforums for providing us the strength to carry on its development. Regrettably, the world is not yet prepared to coexist with a tool of such vast freedom.** Since then there are dozens and dozens of posts selling "WormGPT" which are often ridiculed by users on Telegram and dark market channels who point to the fact the developer openly and loudly shut down the project in August 2023. These appear to be largely jailbreaks or open AI models trained on small phishing email data sets. This has lead to many posts of users claiming wormgpt is a scam, poorly created, or useless though it is often obvious from the user's history that they were using a jailbreak or other scam that was only using the name "wormgpt".