Please Paste This Command to Continue: The ClickFix Trick Aimed at Your Staff

Picture an ordinary Monday. An employee searches for a software download, or clicks a link in an email, and lands on a page that asks them to prove they are human. The page does not ask for a click on traffic lights. It asks the employee to press a keyboard shortcut, paste something, and press Enter. The employee follows the instructions. Nothing looks wrong. The page closes, work continues, and a program built to steal saved passwords is now running on a company computer.

This technique is called ClickFix, and reports published this month show that it is growing and changing in ways that matter to small organizations. It does not exploit a software flaw. It persuades a person to run the attacker's command personally, which is why many email filters and download scanners see nothing suspicious.

What ClickFix is

Microsoft Threat Intelligence describes the pattern in its security blog. A person is sent to a fake page by a phishing email, a malicious advertisement, or a compromised website. The page often poses as a human-verification check, known as a Completely Automated Public Turing test to tell Computers and Humans Apart (CAPTCHA), or as an error that needs fixing. In the background, the page silently copies a command to the computer's clipboard. It then tells the visitor to open the Windows Run box, Windows Terminal, or PowerShell, paste, and press Enter. The pasted command downloads a first-stage program, which then delivers malware such as information stealers or remote access tools.

The key point for a business owner is the last step. The employee, not the attacker, launches the program. Security tools that watch for malicious downloads or attachments may see only a person typing in a normal Windows window.

The numbers this month

Microsoft's 2026 Digital Defense Report, published October 1, 2026, says Microsoft Defender observed ClickFix-style attacker-supplied commands executed on more than 1.1 million unique devices between February and early May 2026. The report describes that as roughly an eightfold increase. The figure comes from Microsoft's own telemetry, so it reflects devices Microsoft can see, but the direction is clear.

A second report adds the newest twist. On October 21, 2026, BleepingComputer published an analysis of ClickFix lures built around artificial intelligence (AI) tools. The article was sponsored by the security vendor Push Security, and the observations in it come largely from that vendor's own detection data, so read it as one vendor's view rather than an industry census. It walks through what it calls InstallFix. A person searches for a popular AI tool and clicks a sponsored search result. The click passes through a compromised website to a fake download page. The page shows what looks like the genuine install command, but its copy button places a different command on the clipboard. According to the article, the malicious command then downloads a script and feeds it straight to the system shell without saving a file to disk.

The same article describes a related tactic: shared conversations and documents on real AI platforms that contain install instructions or links to fake download pages. Because those pages sit on trusted websites, a filter that judges a link by the reputation of its domain may let them through.

Why small businesses are a good target

Three conditions make a small office easier to work with, and none of them is a personal failing.

• Some staff are used to pasting commands. Installing developer and AI tools often does mean copying a line into a terminal. An employee who has done it legitimately once may hesitate less the second time.
• People find software through search. Many employees type the name of a tool into a search engine and click the first result, which may be a paid advertisement.
• Standard accounts may be allowed to run almost anything. Where employees can open the Run box and PowerShell without restriction, a single paste is enough.

The payoff for the attacker can be larger than owners expect. Information stealers are designed to harvest saved browser logins and session data. In our view, the saved logins in one employee's browser can include email, accounting software, the payroll portal, and the cloud storage that holds customer files. That is how a single paste on a single laptop can grow into a company-wide problem.

Six steps that reduce the risk

No single control stops this technique, and the steps below lower the likelihood and the damage rather than remove them. They run roughly from easiest to hardest.

• Set one plain rule for staff. No website, email, chat message, or support caller has a legitimate reason to tell an employee to press a keyboard shortcut and paste a command to prove they are human or to fix an error. If a page asks for that, the employee closes it and tells the owner or the information technology (IT) provider. Microsoft's own recommendation is short: "Ensure users are aware of what they copy and paste." Put the rule in writing and repeat it at a staff meeting. Our guide to verification discipline for social engineering shows how to build the habit.
• Control where software comes from. Decide who may install software, and have that person download tools from the vendor's own website, reached through a saved bookmark or an address typed by hand, rather than from a search advertisement. A sponsored result can look identical to the real one.
• Ask your IT provider about the Run box. Microsoft's mitigation list for ClickFix includes disabling the Run dialog (the Windows key plus R) through Group Policy, and creating an application control policy that blocks Windows programs from launching through Run. Some staff use Run for routine work, so test the change on a few computers first. It is a reasonable use of an hour of your provider's time.
• Add friction and visibility for pasted commands. Microsoft also lists a warning in Windows Terminal when pasted text contains multiple lines, and PowerShell script block logging, which records decoded commands so an investigator can see what ran. Logging does not stop an attack, but it turns a mystery into evidence.
• Turn on the web and email protections you may already own. Microsoft recommends checking Microsoft 365 email filtering for spoofed and malicious mail, rechecking links when they are clicked, and using browsers that support Microsoft Defender SmartScreen, such as Microsoft Edge, along with network protection where your plan includes it. A filter at the Domain Name System (DNS) level can also refuse connections to known-bad sites. We cover that in our protective DNS post.
• Write down what to do after a paste. This step is easy to overlook. If an employee realizes they ran something, the instruction should be to stop working, disconnect the computer from the network, and call the IT provider or the owner right away, without hiding it out of embarrassment. In our view, a fast report is worth more than a perfect one. Treat the passwords and logged-in sessions saved in that browser as exposed. From a different, clean device, change the important passwords and sign out the active sessions, because, as we explain in our session token post, a password change by itself may not end an attacker's access.

A short test for your own office

Ask three questions this week. First, can a standard employee account on your computers open the Run box and PowerShell? Second, if someone pasted a bad command today, would anyone be alerted, and could anyone see what ran? Third, does your team know who to call, and would that person hear about it the same day rather than the next week? If you are unsure about any of these, that is the place to start.

Fake human-verification pages and fake AI-tool downloads will keep changing their appearance. The underlying request stays the same: the attacker needs a person to paste a command. A staff member who has been told that no legitimate website will ask for that is among the more effective defenses against this technique, and it costs a few minutes of a meeting.

Sources

• Microsoft Digital Defense Report 2026, Microsoft, October 1, 2026
• ClickFix has never been bigger, and AI tools are its most convincing lure yet, BleepingComputer (sponsored by Push Security), October 21, 2026
• Think before you Click(Fix): Analyzing the ClickFix social engineering technique, Microsoft Security Blog, Microsoft Threat Intelligence and Microsoft Defender Experts, August 21, 2025