Danger actors are abusing ChatGPT’s content-sharing function to show pretend OpenAI outage pages that direct customers to obtain malware disguised because the ChatGPT desktop software.
The “LLMShare” marketing campaign, found out through Push Safety, makes use of Google advertisements to direct customers looking for ChatGPT to a malicious shared ChatGPT web page hosted on chatgpt.com, permitting the assault to be delivered via a sound OpenAI area.
Customers who click on the commercial are taken to a sound ChatGPT shared web page, however as an alternative of seeing a talk dialog, they’re offered with a rendered outage understand claiming the internet model is unavailable and that they must obtain the desktop software as an alternative.
“We are experiencing prime visitors at this time,” reads the pretend outage message.
“Our web site is briefly unavailable because of a lot of customers. Obtain our desktop app to proceed.”
Not like conventional phishing pages hosted on attacker-controlled infrastructure, the pretend outage understand is rendered via ChatGPT itself.
The attackers created a customized HTML web page the use of ChatGPT’s rendering functions and printed it via a shared chatgpt.com/s/ hyperlink, permitting the pretend outage understand to be displayed from a sound ChatGPT URL.
Push Safety famous that the web page contains “Display code” and “Remix with ChatGPT” controls, revealing that the pretend outage understand is in fact generated from customized HTML and CSS rendered through a ChatGPT steered.
If the customer clicks at the obtain button, they’re delivered to a web site at openew[.]app that impersonates OpenAI’s desktop software obtain portal.
The researchers say the web site makes use of cloaking to show content material simplest to focused sufferers. When safety platforms like URLScan visited the URL, they have been proven a risk free AR/VR corporate web site as an alternative.
The web site gives each macOS [VirusTotal] and Home windows [VirusTotal] downloads that set up malware on units. Whilst it’s unclear what payloads are in the long run deployed, previous campaigns abusing AI platform sharing options have disbursed infostealers.
BleepingComputer’s take a look at of the Home windows model on Any.Run discovered that it executes quite a lot of instructions to resolve whether or not the software is a sound laptop or a digital device.
Push Safety additionally noticed assaults abusing Claude Artifacts, Anthropic’s function for sharing rendered programs and content material, to host ClickFix-style lures that tricked customers into executing malicious instructions.
AI platforms’ sharing options had been abused prior to now to distribute malware to unsuspecting sufferers.
Previous this 12 months, risk actors used Google commercials to direct customers looking for Claude downloads to shared Claude conversations containing malicious set up directions.
Different campaigns abused shared ChatGPT and Grok conversations that performed ClickFix assaults through impersonating device set up guides that advised sufferers to execute instructions that put in malware.
Automatic pentesting equipment ship actual price, however they have been constructed to respond to one query: can an attacker transfer during the community? They weren’t constructed to check whether or not your controls block threats, your detection regulations hearth, or your cloud configs grasp.
This information covers the 6 surfaces you in fact want to validate.
Obtain Now
