Ransomware Operators Exploit the AI Hype
As reported by Bleeping Computer, cybercriminals are increasingly using fake AI tools to deliver ransomware and other malicious payloads, according to new findings from Cisco Talos. This tactic builds on a trend that began last year when advanced threat actors first began using deepfake content and AI-themed lures to infect victims. Now, the approach has expanded to include a wider range of threat actors, including smaller ransomware groups and info-stealer operators.
Cisco researchers have identified multiple campaigns that impersonate legitimate AI tools through deceptive websites, SEO poisoning, and malvertising, all designed to lure victims into downloading malware. These campaigns aim to breach corporate networks by exploiting the public’s growing interest in artificial intelligence.
CyberLock Delivered Through Fake AI Website
One of the groups identified, known as CyberLock, is distributing PowerShell-based ransomware via a website impersonating the legitimate novaleads.app platform. The malicious site, novaleadsai[.]com, offers a fake “free 12-month subscription” to an AI service. Victims who download the installer receive a .NET loader that deploys the ransomware.
Once executed, CyberLock encrypts files across multiple disk partitions and appends a .cyberlock extension. The ransom note demands $50,000 in Monero and claims the payment will be used to support humanitarian causes in regions including Palestine, Ukraine, Africa, and Asia. The malware also replaces the victim’s desktop wallpaper with an image referencing a SentinelLabs blog, further signaling compromise.
Lucky_Gh0$t Disguised as ChatGPT Installer
A second campaign observed by Cisco involves Lucky_Gh0$t, a ransomware variant based on the Chaos-derived Yashma family. It is distributed as a fake installer titled “ChatGPT 4.0 full version – Premium.exe” within a self-extracting archive. The archive includes legitimate Microsoft open-source AI tools alongside the ransomware payload, likely to avoid antivirus detection.
When executed, Lucky_Gh0$t encrypts files smaller than 1.2 GB and appends a random four-character extension. Larger files are overwritten with junk data and deleted. Victims are provided a personal ID and instructed to contact the attackers via the secure messaging platform Session to begin ransom negotiations.
Numero Malware Disrupts Systems Without Encryption
In addition to ransomware, Cisco Talos also uncovered a new malware strain named Numero. Posing as an installer for the video editing platform InVideo AI, the dropper includes a batch file, VBScript, and an executable named wintitle.exe. Once launched, Numero enters an infinite loop that corrupts the graphical user interface by replacing all visible text with the numeric string “1234567890.”
Unlike traditional ransomware, Numero does not encrypt data or issue ransom demands. However, it renders the system unusable by visually locking it in a corrupted state, effectively requiring a system reinstallation to recover.
Malvertising and SEO Poisoning Drive Distribution
All three threats rely heavily on malvertising and search engine optimization (SEO) poisoning to reach victims. By manipulating search engine results and purchasing ad placements, attackers are able to position malicious downloads prominently in search queries related to AI tools.
As public interest in artificial intelligence continues to grow, so too will cybercriminals’ attempts to exploit it. These campaigns demonstrate how quickly threat actors can adapt to emerging trends in technology to improve their social engineering and delivery methods.
Security Recommendations
Security experts advise users to be cautious when downloading AI tools or any software from unfamiliar websites. Official software should always be obtained directly from verified sources rather than through promoted links, social media posts, or third-party websites.
In addition to securing download practices, organizations should implement structured security awareness programs that include training on recognizing SEO poisoning, social engineering tactics, and indicators of malicious software. Technical teams should reinforce these efforts with internal communication about emerging threats to keep users informed.
