ClickFix now leads phishing detections, and attackers are using AI tool brands as bait
Fake install pages for Claude Code, Codex and NotebookLM show genuine-looking commands while copying malicious ones, a trick that exploits trust in the AI brands themselves.
ClickFix, a social-engineering technique that tricks users into pasting malicious commands, now accounts for an average of 52% of detections at Push, overtaking AiTM and device code phishing for the first time, and reached 67% in August. Criminals increasingly use AI tools as bait: InstallFix pages clone real install pages for Claude Code, Codex and NotebookLM and show Anthropic's genuine command while copying a different one to the clipboard. ANY.RUN found that 81% of 1,258 fake AI-tool sandbox analyses from April 15 to October 5, 2026 impersonated Claude, ChatGPT or Codex. Attackers also abuse AI chat sharing and custom GPTs, which are trusted because they are hosted on chatgpt.com or claude.ai. Microsoft reports ClickFix attacks grew roughly 23x between December 2025 and May 2026.
Why it matters
The lure works because users trust the source they arrive from, not because the technique is new. When a fake page sits under a familiar AI brand, the user's own habit of copying a setup command becomes the attack path, and that habit is hard to train away. The growth in ClickFix detections over a few months suggests this is a shift in where attackers are putting their effort, away from older methods and toward impersonating tools people already rely on.
Matters or hype?
Does this story matter, or is it hype? Vote before you see what everyone else said.
Sources
- BleepingComputer