186 TypoSquatted AI Domains Deliver ClickFix Lures, Net Guard Malware, and Search Hijacking Extensions

Cybersecurity researchers at Island have identified a large-scale campaign using 186 lookalike domains that impersonate popular AI tools from 27 brands. The operation, named TypoPrompting, routes visitors to ClickFix-style social engineering pages, the Net Guard browser hijacker, or malicious Chrome extensions that override search providers. Domains mimicking Anthropic Claude, OpenAI ChatGPT, Google Gemini, Microsoft Copilot, and others have been active since at least July 2023. One path leads through fingerprinting scripts to fake Cloudflare verification pages that execute PowerShell commands, while another delivers the Net Guard MSIX package capable of injecting JavaScript and altering browser settings. A third vector promotes 16 extensions, including Quick Search, that collectively claim nearly one million users and redirect queries to third-party services.

The Hacker NewsTranslated from English

Read at source

Cybersecurity researchers have discovered that 186 bogus domains impersonating artificial intelligence tools are being used to distribute ClickFix lures, browser-hijacking malware, and search-manipulating extensions. The campaign has been named TypoPrompting by the company Island.

The malicious domains target 27 well-known AI brands, including Anthropic Claude, OpenAI ChatGPT, Google Gemini, Microsoft Copilot, SpaceXAI Grok, DeepSeek, Mistral, Perplexity, Midjourney, ElevenLabs, Lovable, and Cursor. Examples of registered lookalikes include cloude[.]ai, cluade[.]ai, clsude[.]ai, chatlgpt[.]com, geimini[.]com, notebookklm[.]com, griok[.]com, preplexity[.]ai, deepceek[.]com, elevenlebs[.]io, copiot[.]com, and midj0urney[.]org. Registration records show that some of these domains, such as cloude[.]ai and cluade[.]ai, were created as early as July 13, 2023.

Three distinct infection paths

The first route presents a fake Cloudflare verification page that runs a clipboard-based PowerShell command after browser fingerprinting and IP collection. The second route displays a “Security Check” page that ultimately installs Net Guard, an application that claims to be a legitimate Wi-Fi and DNS scanner but contains capabilities to execute attacker-controlled JavaScript, modify the default search engine, edit browser profiles, and install arbitrary extensions. The third route redirects users to the Quick Search extension and 15 additional listings that override the browser’s search provider with find.flashysearches[.]com, sending queries—including potentially sensitive internal hostnames—to third parties.

The fingerprinting script used in the ClickFix path is shared across 211 entry domains that also impersonate banks, insurers, and enterprise platforms such as Scotiabank, Charles Schwab, Blue Cross Blue Shield, Verizon, ServiceNow, and Microsoft Azure. Eight additional typosquatted domains specifically targeting Claude, DeepSeek, Copilot, ElevenLabs, and Voice AI lead to the Net Guard distribution pages.

Island researchers noted that the 16 malicious extensions currently listed on the Chrome Web Store have accumulated 972,000 declared users as of November 3, 2026, with one extension alone claiming 80,000 users. The company emphasized that the increasing integration of AI services into daily workflows makes accurate domain resolution a critical security control rather than a matter of typing accuracy.

Meridian digestThe week's most consequential security stories, distilled into one briefing.See the digest

More in Malware & Botnets

Europol and Partners Seize 257 Domains and 235 Servers in Major Operation Endgame Strike Against Gootloader, CastleRAT, and StealC
Threat Actors Deploy SCKit Malware Toolkit Through Compromised ConnectWise ScreenConnect Panels
Threat Actors Abuse Legitimate GSocket Tool for Stealthy Persistent Access on Linux Servers