• A QR code is a link in disguise. Check where it goes before trusting where it came from.
• Context is your best detector. A sticker over a code, a menu that wants your password, or a parking sign that wants your bank login is a mismatch worth a pause.
• Report it, even if you’re not sure. A photo and a quick note to IT/security can save the next person.

Previously, in Part Two: an agent that treated “no” as a puzzle, a state-backed crew walking in on a vendor’s stolen keys, a ransomware freelancer running break-ins by chatting with his coding assistant, and defenders getting robots of their own. Then we looked ahead: agents getting ID badges, shadow AI growing up, and patching becoming a survival skill. That was AI that acts.

This finale is about something smaller and sneakier: the little square that can send you somewhere bad, and how to not get fooled by it.

Imagine arriving at a conference with a coffee in one hand and approximately four minutes to find your session. A handsome sign announces: “Updated agenda. Scan here.” It has the right logo, the right colors, and the reassuring confidence of something that has been laminated.

You scan. A page asks you to sign in with your work email. Fair enough, you think. Conferences have apps. Apps have accounts. Everything apparently needs an account now, including the toaster.

Except, in this imagined example, someone planted the sign. You are about to give your password to a stranger because the stranger owns a printer.

That is the trick we need to understand. A professional-looking sign tells you very little about the website behind its QR code. Fortunately, you can check more useful things than the quality of the lamination.

What that little square actually does

QR stands for Quick Response. A QR code is a barcode your phone can read. It can hold a web address, payment information, or instructions such as joining a wireless network.  For the everyday website codes discussed here, the useful takeaway is simple: a QR code is a link you cannot inspect at a glance.

Phishing means impersonating someone trustworthy to trick you into handing over information, money, or access. QR-code phishing uses a code to deliver that deception. You may hear it called quishing, because apparently the security industry decided the scam also needed an unfortunate nickname. A fraudulent destination might imitate a bank, a parking service, or your employer’s sign-in page. It might also offer malware, meaning software designed to harm a device or steal information.

Reading a code, opening its destination, entering information, and installing software are different actions. Apple’s current instructions, for example, describe scanning the code and then tapping the displayed link to open it. Scanner behavior varies, so use the preview when available. A scan by itself does not establish that your phone or account has been compromised.

Opening an unfamiliar page still carries risk. Keep your phone updated and stop if anything seems wrong.

The danger often comes when the next request feels ordinary enough that we stop questioning it.

Why criminals have fallen for a little square

Criminals like QR codes for the same reason the rest of us do: they’re quick. But crooks enjoy two extra perks. First, a code hides its link from the tools built to catch bad links. Email filters, the software that screens incoming mail for scams, are excellent at reading text, and a QR code is a picture. Second, the scan usually happens on a phone, which carries the whole transaction away from the protections on your work laptop and onto a small screen where addresses are easy to misread.

The enthusiasm shows in the numbers. Microsoft reported that the QR-code phishing it detected rose 146% across the first three months of 2026. Security firm ESET found QR codes in roughly 11% of all the phishing emails it saw in the first half of the year.

And the crooks keep redecorating. Criminals now rent ready-made scam toolkits from one another, a business model called phishing-as-a-service. One such kit splits a QR code into two separate images that only work as a code when your phone sees them side by side; another wraps a malicious code around a harmless one so automated scanners get a muddled answer. Every one of those tricks is aimed at fooling machines. None of them changes what happens after you scan, and that’s the part where you get to win.

Researchers have tested the problem

A study published at USENIX Security in August 2025 examined QR-code attacks across three experiments involving 1,876 participants. In its fraudulent-payment comparison, 13% of participants using QR codes recognized the fraud, compared with 46% of those entering payment details manually. That comparison involved 443 participants.

These were controlled experiments conducted in Germany. The authors caution that their results do not measure everyone’s real-world chances of being scammed. The practical lesson is that removing a little typing can also remove an opportunity to notice that the recipient or payment details are wrong.

There is encouraging research, too. A research paper published in April 2026 found that redesigned scanner warnings and a short tutorial improved participants’ ability to distinguish legitimate from deceptive destinations in simulated tests. This does not establish how well an app stops attacks in everyday use, but it gives us a useful direction: make the destination clearer and help people understand the decision.

We can work with that. Nobody needs to learn to read the black squares by eye.

The conference agenda that wants your password

The opening scene was fictional. Conference-themed attacks are documented.

In a January 8, 2026 alert, the FBI described a June 2025 attack in which an invitation to a nonexistent conference contained a QR code. It opened a registration page whose registration button led to a fake Google sign-in page designed to capture login details.

On October 8, 2026, Cisco Talos reported a campaign targeting people affiliated with research organizations in Taiwan. Attackers copied legitimate event-poster designs, altered the QR codes, and distributed the posters as email attachments. The registration deception led toward fake Google sign-in pages. Talos suggested that the posters might also be printed and displayed, but did not establish that this happened.

Talos found something nastier under the hood: an adversary-in-the-middle (AitM) kit. Instead of simply copying your password, the fake page sits between you and the real Google sign-in, passing along everything you type, including the one-time code from your phone, as you type it. That’s how it slips past multifactor authentication, the extra login check most of us rely on. Spoiler: passkeys, which come up in the habits below, are built to shut this trick down.

That is a particularly useful lesson: real event details can surround a fraudulent destination. The speaker may exist. The venue may exist. The date may be correct. None of those facts gives the person operating the linked website permission to have your password.

Consider the same trick applied to a bogus agenda sheet or an invented “Express registration check-in” sign beside a queue. Those are illustrative scenarios, but the decision is practical: open the event app you already trust or ask an identifiable registration staff member to confirm the process. Some legitimate events use outside registration providers, so an unfamiliar address needs verification rather than an automatic guilty verdict.

For an organization behind one of the world’s best-known cybersecurity conferences, take this as backhanded flattery: attackers copy events because people trust events. That trust is worth protecting, starting with a two-second look at the address.

The restaurant menu with unusual ambitions

Now imagine a restaurant menu code covered by a replacement sticker. You expect appetizers; the page asks you to install a “menu viewer,” enter your email password, or provide card details before showing the menu.

The soup has developed some remarkable administrative requirements.

Treat that mismatch as a reason to stop and ask your server. Ordering and paying legitimately require more information than simply reading a menu. Verify which task you are actually doing. A sticker is a clue to check, not proof of fraud; businesses sometimes update their codes legitimately.

Keep the risk in perspective. The UK’s National Cyber Security Centre says restaurant and pub QR codes are generally likely to be safe, while publicly accessible codes in places such as car parks can be riskier. We can enjoy lunch while retaining the right to ask why the lunch menu wants access to our account.

The parking notice that sends your money elsewhere

Parking is an excellent setting for this scam because the request fits the moment. You need to pay, you may be late, and nobody wants to conduct a forensic investigation beside a parking meter.

This one keeps showing up in the news. In September 2026, the Federal Trade Commission warned again about scammers covering legitimate parking codes with their own. The same month, Irish police in Donegal warned that fake stickers were turning up on EV charging stations as well as parking meters. In October, police in Ann Arbor, Michigan, found stickers reading “Scan to Park!” and “Park AA” on downtown meters, and city employees were sent out to inspect meters and peel off anything suspicious. A sheet of stickers costs pennies. A city worker with a fingernail costs the scammer the entire investment.

Use the parking operator’s verified app, a known website, or the meter’s normal payment method. If a notice claims that the machine is broken and only its QR code works, verify that instruction independently. Before confirming any payment, check the recipient, amount, and parking location. A payment screen can look wonderfully efficient while efficiently paying the wrong person.

The free gift with a surprisingly expensive catch

Imagine a fake promotional flyer promising a prize, a conference giveaway, or a discount. After scanning, you are asked for a small shipping payment, a password, or more personal information than the offer reasonably needs. The prize supplies the distraction; the next request supplies the opportunity.

A related real-world warning came from the FBI on July 31, 2025. It described unsolicited packages containing QR codes that draw recipients toward requests for personal or financial information or malicious software downloads. Missing sender information can encourage people to scan out of curiosity. The FBI explicitly said this variation was less widespread than other fraud schemes.

Verify an offer through the company’s established website or a representative you can independently identify. “Free” is an appealing price. It becomes less appealing when the checkout includes your email account.

Coming up in Part 3 (cont.): the QR codes that don’t send you anywhere at all, how to read an address in two seconds, nine habits worth keeping, and what to do if you’ve already scanned something suspicious.