Affected Systems
Organizations using AI tools (Claude, ChatGPT, Codex, NotebookLM). All platforms (Windows, macOS). Users accessing AI tools via search engines or shared chatbot links. ClickFix now represents 52-67% of browser-based attacks detected in Q2-Q3 2026.
Exploitation Status
Active widespread exploitation. Push Security reports ClickFix reached 67% of detections in August 2026. ANY.RUN identified 1,258 sandbox analyses linked to fake AI campaigns between April-October 2026. Malicious custom GPT observed with over 10,000 conversations. Attacks delivered via malvertising, SEO poisoning, compromised websites, and shared chatbot links.
Business Impact
High risk of initial access and malware deployment. 80% of ClickFix attacks in 2026 originated from search engines, bypassing email security controls. Attacks exploit legitimate user behavior (terminal installs, AI tool searches) and evade URL reputation filters by hosting lures on trusted domains (chatgpt.com, claude.ai). Cloaking techniques prevent detection by security tools. Targets employees adopting unapproved AI tools faster than security teams can vet them.
Urgency
🟠Within 24 hours
Recommended Actions
- Block execution of commands pasted from clipboard in terminal environments using endpoint controls or EDR policies where feasible
- Deploy browser security solutions capable of intercepting ClickFix techniques including clipboard manipulation and script injection (e.g., Push Security, Menlo Security)
- Implement DNS filtering and search ad blocking to prevent access to malvertising and SEO-poisoned results for high-risk keywords (AI tool names)
- Establish approved AI tool catalog with direct URL bookmarks; educate users to never search for AI tools via search engines or install via unverified commands
- Monitor for anomalous PowerShell/bash execution patterns and outbound connections following clipboard paste events; alert on encoded command execution piped to shell