From 6bd00a475cfb2fddf169e37be836ba62182272c0 Mon Sep 17 00:00:00 2001 From: nsosio Date: Tue, 29 Aug 2023 20:31:00 +0200 Subject: [PATCH] first draft of FraudGPT; included links for WormGPT, Darkbert and PoisonGPT --- uncensored-models.md | 46 ++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 44 insertions(+), 2 deletions(-) diff --git a/uncensored-models.md b/uncensored-models.md index d4bf251..bdd2b86 100644 --- a/uncensored-models.md +++ b/uncensored-models.md @@ -1,7 +1,49 @@ # Uncensored Models -FraudGPT, PoisonGPT +## FraudGPT -"Cybercriminals train AI chatbots for phishing, malware attacks" (WormGPT: ChatGPT clone trained on malware-focused data, new: FraudGPT, coming soon: Bard-based version) https://www.bleepingcomputer.com/news/security/cybercriminals-train-ai-chatbots-for-phishing-malware-attacks +FraudGPT has surfaced as a concerning AI-driven cybersecurity anomaly. Operating in the shadows of the dark web and platforms +like Telegram, FraudGPT mimics ChatGPT but with a sinister twist—fueling cyberattacks. Unlike its legitimate counterpart, +FraudGPT sidesteps safeguards, making it a playground for malicious content. + +[Netenrich](https://netenrich.com)'s vigilant threat research team unveiled FraudGPT's ominous potential in July 2023. The subscription fee +for FraudGPT starts at $200 per month and goes up to $1,700 per year (as stated by +[Netenrich's vigilant threat research team](https://netenrich.com/blog/fraudgpt-the-villain-avatar-of-chatgpt)). +The tool's interface mirrors ChatGPT, allowing users to prompt responses tailored for nefarious +deeds. + +Test scenarios involved crafting bank-related phishing emails, with guidance on +embedding malicious links for impact. The ability to create scam landing pages and identify targets highlights its +expedited malicious campaign potential. + +Amid evolving cybercriminal tactics, FraudGPT serves as a stark reminder of the shifting threat landscape. Its rise +underscores hackers' inclination to exploit cutting-edge tools. In this ever-evolving scenario, cybersecurity professionals +and individuals must remain adaptable and prepared to counter such emerging threats. + +Links +- https://hackernoon.com/what-is-fraudgpt +- https://slashnext.com/blog/ai-based-cybercrime-tools-wormgpt-and-fraudgpt-could-be-the-tip-of-the-iceberg/ + +## WormGPT + +Links +- https://slashnext.com/blog/wormgpt-the-generative-ai-tool-cybercriminals-are-using-to-launch-business-email-compromise-attacks/ +- https://www.geeksforgeeks.org/wormgpt-alternatives/ +- https://slashnext.com/blog/ai-based-cybercrime-tools-wormgpt-and-fraudgpt-could-be-the-tip-of-the-iceberg/ + +## Darkbert + +Links +- https://www.esecurityplanet.com/threats/fraudgpt-darkbert-darkbart/ + +## PoisonGPT + +Links +- https://blog.mithrilsecurity.io/poisongpt-how-we-hid-a-lobotomized-llm-on-hugging-face-to-spread-fake-news/ +- https://aitoolmall.com/news/what-is-poisongpt/ + + +"Cybercriminals train AI chatbots for phishing, malware attacks" (WormGPT: ChatGPT clone trained on malware-focused +data, new: FraudGPT, coming soon: Bard-based version) https://www.bleepingcomputer.com/news/security/cybercriminals-train-ai-chatbots-for-phishing-malware-attacks {{ comments }}