Fake Claude AI Installer on Google Sites Delivers Credential-Stealing Malware

Key Takeaways

• Threat actors are distributing a malicious installer for “Claude AI” through Google Sites, designed to steal user credentials.
• The fake installer deploys Lumma Stealer, an info-stealing malware capable of exfiltrating sensitive data from browsers, cryptocurrency wallets, and system information.
• The campaign leverages sophisticated social engineering and legitimate services to appear credible, increasing the likelihood of user compromise.
• Users should exercise extreme caution when downloading software, especially AI applications, and verify sources directly from official vendor websites.

Malicious “Claude AI” Installer Discovered on Google Sites Delivering Credential Stealer

Cybersecurity researchers have uncovered a new campaign distributing a sophisticated information-stealing malware, Lumma Stealer, disguised as an installer for Anthropic’s Claude AI. The malicious files are being hosted on Google Sites, exploiting the platform’s legitimacy to appear trustworthy to unsuspecting users.

The deceptive operation aims to trick individuals searching for the popular artificial intelligence chatbot into downloading a compromised application. Instead of gaining access to Claude AI, victims inadvertently install malware designed to exfiltrate a wide array of sensitive personal data.

How the Attack Unfolds

The attack vector begins with threat actors creating fake download pages on Google Sites. These pages are meticulously crafted to mimic legitimate software distribution sites, featuring compelling calls to action for downloading what purports to be the Claude AI desktop application.

Once a user clicks the “download” button, they are prompted to retrieve an executable file. This file, however, is not the benevolent AI assistant but a dropper for Lumma Stealer. Lumma Stealer is a known commodity in the cybercrime underground, recognized for its efficiency in harvesting credentials and other valuable information.

Lumma Stealer: A Potent Info-Stealer

Lumma Stealer operates by targeting critical user data stored on compromised systems. Its capabilities include, but are not limited to:

• Extracting saved passwords, cookies, and autofill data from web browsers such as Chrome, Firefox, Edge, and Brave.
• Harvesting cryptocurrency wallet information, potentially leading to the theft of digital assets.
• Collecting system information, including IP addresses, operating system details, and hardware configurations, which can be used for further targeted attacks.
• Stealing data from various desktop applications and messaging clients.

The malware typically communicates with a command-and-control (C2) server to exfiltrate the stolen data, often using encrypted channels to evade detection.

Leveraging Legitimate Infrastructure for Malicious Ends

The choice to host the malicious installers on Google Sites is a deliberate tactic by the attackers. Google Sites, being a legitimate and widely trusted platform, often bypasses initial security checks by users and some automated security solutions. This lends an air of authenticity to the fake download pages, making them more effective in social engineering efforts.

This campaign highlights a growing trend where cybercriminals abuse trusted cloud services and platforms to host their malicious infrastructure, complicating detection and takedown efforts. The perceived legitimacy of these platforms makes it harder for average users to discern genuine software from fraudulent offerings.

The Broader Threat Landscape of AI-Themed Scams

The proliferation of AI technologies has unfortunately created fertile ground for new cyber scams. Threat actors are quick to capitalize on public interest and demand for AI tools, crafting convincing lures that promise early access or enhanced features for popular AI models like Claude AI, ChatGPT, and others. Users should remain vigilant, as these scams often prey on the excitement surrounding new technological advancements.

What You Should Do

• Verify Software Sources: Always download software directly from the official vendor’s website. Avoid third-party download sites, even if they appear legitimate. For Claude AI, access should be through Anthropic’s official channels.
• Exercise Caution with Links: Be suspicious of unsolicited links or downloads, especially those promoted through social media or unfamiliar websites.
• Use Antivirus/Endpoint Protection: Ensure your antivirus software or endpoint detection and response (EDR) solution is up-to-date and actively scanning for threats.
• Enable Multi-Factor Authentication (MFA): Implement MFA on all accounts, especially those containing sensitive information like email, banking, and cryptocurrency wallets. This adds an extra layer of security even if credentials are stolen.
• Educate Yourself: Stay informed about common social engineering tactics and the latest cybersecurity threats.

Disclaimer: HackersRadar reports on cybersecurity threats and incidents for informational and awareness purposes only. We do not engage in hacking activities, data exfiltration, or the hosting or distribution of stolen or leaked information. All content is based on publicly available sources.