186 Fake AI Domains Redirect Users to ClickFix Lures, Browser Malware, and Search Hijackers

The cybersecurity company Island has identified a widespread campaign using 186 bogus domains that mimic popular AI tools. These lookalike domains serve three types of threats: ClickFix-style lures, browser-hijacking malware called Net Guard, and search-hijacking extensions. The campaign spans 27 AI brands including Anthropic Claude, OpenAI ChatGPT, Google Gemini, Microsoft Copilot, and others.

The domains use subtle misspellings or variations to trick users into believing they're accessing legitimate AI services. Examples include cloude[.]ai, chatlgpt[.]com, geimini[.]com, and preplexity[.]ai. Some of these domains were registered as early as July 13, 2023, indicating a long-term operation.

Three Threat Pathways

The first category of fake domains leads to ClickFix lures that serve PowerShell commands through fake Cloudflare verification pages. The second route directs users to "Security Check" pages that deliver Net Guard malware, which can run attacker-controlled JavaScript and change browser settings. The third pathway promotes search-hijacking extensions available on the Chrome Web Store.

Browser Extension Hijacking

The most concerning aspect involves browser extensions like Quick Search, available on the Chrome Web Store. These extensions override default search providers with find.flashysearches[.]com, potentially exposing sensitive workplace information including internal hostnames, customer names, and project details. The Quick Search extension is part of a larger set of 16 similar extensions with a combined 972,000 users as of November 3, 2026.

The extensions require minimal permissions to operate, making them particularly dangerous. "The analyzed extension needs no executable JavaScript or declared permissions to perform its main action," Island noted, highlighting how easily users can be compromised.

Broader TypoPrompting Campaign

This AI-focused campaign is part of a larger scheme called TypoPrompting that includes lookalike domains of banks, insurers, and enterprise software. These domains serve fake verification pages, terminal updates, and archive downloads designed to execute malicious commands on both Windows and macOS systems.

The campaign exploits parked domains that redirect visitors to scam and malware sites. When a domain owner opts into direct search, traffic to the domain is sold to advertisers who bid on keywords and traffic characteristics, according to DNS threat intelligence firm Infoblox, which highlighted similar tactics in December 2025.