Any Captcha that asks you to press keyboard shortcuts is a scam

A Captcha that tells you to press a set of keys and then Enter is not a security check. It gets you to run a hidden command that installs malicious.

A Captcha that tells you to press a set of keys and then Enter is not a security check. It gets you to run a hidden command that installs malicious software. No genuine human-verification test works this way.

Key takeaways

  • A genuine Captcha works entirely inside the web page and never asks you to use keyboard shortcuts that control your operating system.
  • In the key-press scam, the page quietly copies a command to your clipboard and the instructions get you to paste and run it yourself.
  • Because the victim runs the command, the attack can get past some of the warnings that normally appear when you download a suspicious file.
  • A website you trust can still show a fake Captcha if criminals have compromised it, so a good reputation is no protection.

The argument: the instructions themselves are the warning sign

The Guardian reports on a scam built around a fake “prove you are not a robot” test. In the scenario it describes, someone shopping for Christmas presents visits the website of a small business they have heard good things about. A verification prompt appears. It does not ask them to pick out images or type distorted letters. It asks them to press a series of keys and then Enter. The Guardian says this is how fraudsters get malicious software onto the computer.

The argument of this article is simple. Asking someone to press keyboard shortcuts is enough on its own to show that a verification prompt is fraudulent. You do not need to judge whether the site looks professional, whether the design seems right or whether the business has a good name. If a “human check” asks you to open something on your computer, paste into it and press Enter, it is an attack. You can close the page.

A simple rule like this matters because the scam relies on habit. People have spent years clicking through Captchas without thinking, so a new kind of Captcha looks like an inconvenience rather than a threat. The scam turns that familiarity against them. A clear rule that needs no technical knowledge breaks the habit.

Evidence one: real Captchas live inside the browser

Legitimate Captcha systems are designed to work entirely within the web page. In their familiar forms they ask you to tick a box, choose the pictures that contain a certain object, type out distorted text or solve a small puzzle. Many now run in the background with no visible challenge, judging whether a visitor is human from signals the page can gather for itself.

All of these designs share one feature: the test runs inside the browser and the answer goes back to the website. Proving you are human never requires you to step outside the browser. A website has no legitimate reason to ask you to open a system dialog box, a command prompt or a terminal to show that you are human. Those tools control your computer, not the web page, and they play no part in how a real Captcha confirms that a visitor is a person.

That is why the key-press request is such a reliable sign. It is not a slightly odd version of a normal test. It asks for something no genuine test needs. When a prompt asks you to use keyboard combinations that open parts of your operating system, it has moved from verifying you to taking control of your machine.

Evidence two: the victim is made to run the attack

Security researchers have described this kind of attack in detail, and some refer to it as “ClickFix”. The general pattern on Windows computers is as follows.

When the visitor clicks the fake verification box, the page uses the browser’s clipboard feature to copy a line of text without saying so. That text is a command, often one that tells a built-in Windows scripting tool such as PowerShell to fetch and run a program from a remote server. The on-screen instructions then tell the visitor to press the Windows key and R together, which opens the Run dialog box. Next they press Ctrl and V, which pastes the hidden command into that box. Finally they press Enter, which runs it.

At no point does the visitor see a file download in the usual way. Each step looks harmless and is presented as part of a routine check. Taken together, the steps give the attacker’s command to the operating system with the user’s own authority. Researchers have reported variants aimed at other systems, such as instructions on Apple computers that lead the user to open the Terminal application and paste a command there.

This design explains why the attack works. Browsers and operating systems have many defences against software that tries to install itself: download warnings, file-reputation checks and prompts when an unknown program first runs. A command the user types or pastes into a system tool can avoid some of those checkpoints, because the system treats it as something the user deliberately chose to do. The fraudsters do not have to break through the computer’s defences. They persuade the owner to open the door.

What happens next depends on the attacker. Malicious software installed this way can include programs that steal saved passwords, browser cookies or financial details, or that give criminals remote access. The Guardian’s account does not say which specific malicious programs are being delivered in the cases it describes, and that detail is not known here.

Evidence three: trusted websites can carry the trap

The Guardian’s example is worth noting because the victim does nothing obviously risky. They are not opening a strange email attachment or visiting a dubious download site. They are going to a small business website they have reason to trust.

This matches how researchers describe the delivery of fake Captchas more generally. Criminals can plant the prompt on legitimate websites they have compromised, often by exploiting weaknesses in the software those sites run on. They can also deliver it through malicious adverts, links in phishing messages or lookalike pages that copy real brands. In each case, the reputation of the site in your address bar tells you little about whether the page in front of you has been tampered with.

Small businesses can be especially exposed. They often lack dedicated security staff and may not notice for some time that their site has been altered. A visitor in the Guardian’s scenario has no realistic way to tell whether the site has been compromised. What they can recognise is the unusual request. This is why the warning sign has to be the instruction itself, not the site that delivers it.

The case against: a simple rule is not enough on its own

The strongest objection is that a single rule puts too much weight on individual users and too little on the systems around them. Several points support this view.

First, attackers adapt. If the public learns that “a Captcha asking for key presses is a scam”, criminals can change the pretext. The same clipboard technique can be presented as a fix for a supposed browser error, a step needed to view a document or an instruction to join a video call. Someone who has learned a rule specifically about Captchas may not recognise the same attack in different clothing. Any rule tied to one disguise has a short shelf life.

Second, people make mistakes under pressure. Tiredness, distraction or eagerness to finish a purchase can lead someone to follow instructions they would normally question. A defence that depends on perfect attention every time will sometimes fail.

Third, much of the risk can be reduced by technical measures that do not depend on the user at all. Organisations can restrict who is allowed to use scripting tools or the Run dialog on managed computers. Security software can watch for suspicious commands that fetch code from the internet. Browser makers can look again at how freely web pages are allowed to write to the clipboard. Website owners can keep their software up to date so their pages are harder to compromise. Supporters of this view argue that these measures, rather than public awareness, are what will eventually make the attack unprofitable.

Finally, focusing on the victim’s actions can slide into blaming people for being deceived by a scheme designed by professionals to deceive them.

These points are fair. They do not, however, undermine the core claim. The rule is not offered as the only defence. It is offered as the one defence any individual can use immediately, without special software or knowledge. The broader principle behind it holds whatever the disguise: no legitimate website needs you to paste and run commands on your computer. Stated that way, the rule covers the Captcha version and the variants alike.

Evidence that would change this conclusion

Several developments would weaken the argument made here.

If legitimate verification services began asking users to press keyboard combinations, perhaps as an accessibility option for people who cannot use a mouse or see images, the key-press request would no longer reliably signal fraud. The rule would then need to be narrowed to the specific act of opening system tools or pasting commands. There is no indication that this is happening, but it would change the advice.

If studies showed that awareness campaigns about this scam made little difference to infection rates, the case for leading with a user-facing rule would weaken. Effort would be better spent on technical controls. Reliable public data on how often the attack succeeds, and whether warnings reduce that rate, is not available here.

If operating system and browser makers changed how pasted commands are handled, for example by adding clear warnings before running commands that download code from the internet, or by limiting what web pages can silently place on the clipboard, the attack in its current form could become far less effective. The rule would remain sensible, but it would matter less.

Finally, if attackers moved to methods that need no user action at all, the focus of public advice would have to shift. For now, the defining feature of this scam is that it needs the victim’s own hands on the keyboard. That is its weakness, and it is why recognising the request is such an effective defence.

Sources and further reading

  • The Guardian: its consumer and money coverage describing the fake Captcha key-press scam and the scenario of a shopper visiting a small business website.
  • Security research and threat intelligence publications from major technology and cybersecurity companies: technical write-ups describing the clipboard-based technique known as ClickFix.
  • National cybersecurity agencies: general public guidance on phishing, malicious websites and safe browsing practice.
  • Documentation from Captcha and bot-detection providers: explanations of how legitimate human-verification systems work within the browser.

Surfaced from the rss:guardian_tech signal “fake Captcha malware scam”. AI-assisted draft, editorially reviewed.

Visited 3 times, 1 visit(s) today
share this recipe:
Facebook
X
WhatsApp
Telegram
Email
Reddit