A small pixelated square. You’ve scanned hundreds of them at coffee shops, parking meters, airport lounges, and restaurant tables. It feels instant. It feels safe. It feels routine. That’s exactly what cybercriminals are counting on.
QR codes have become one of the most trusted and least scrutinized gateways in everyday digital life. A new category of attack known as “quishing” (QR + phishing) is now one of the fastest-growing cyber threats globally, silently bypassing some of the most sophisticated enterprise email security systems in the world.
Between August and November 2025, detections of malicious QR code phishing emails surged fivefold from 46,969 incidents in August to 249,723 in November alone. Over 4.2 million QR code phishing threats were identified in the first half of 2025. Microsoft flagged more than 15,000 daily QR-code-bearing phishing emails targeting the education sector alone.
Quishing works by embedding a malicious URL inside a QR code image. When a victim scans it, their phone camera decodes the hidden URL and the browser automatically opens the malicious destination a fake website designed to steal credentials, payment card details, or install malware.
Unlike a phishing link in an email where you can hover to preview the URL, a QR code reveals nothing until after you’ve scanned it. The malicious payload is locked inside an image, completely invisible to the naked eye and, critically, unreadable by most email security filters.
Traditional email security gateways were designed to inspect text, parse HTML, and evaluate embedded URLs. When a malicious payload is encoded inside the pixel matrix of an image, the gateway sees only a JPEG or PNG file; it finds no suspicious links and delivers the message to the inbox. This is not a configuration failure. It is an architectural limitation.
As a result:
| Metric | Statistic | Source |
|---|---|---|
| QR phishing emails (Aug–Nov 2025) | 46,969 → 249,723 (5× surge) | Kaspersky |
| QR phishing threats identified (early 2025) | Over 4.2 million | Keepnet Labs |
| QR code phishing surge (2026) | +146% year-over-year | Microsoft |
| % of phishing attacks using QR codes (2025) | 12% | Keepnet / Linkcpa |
| % of QR attacks targeting mobile users | 68% | Keepnet Labs |
| % of QR attacks aimed at credential theft | ~89–90% | Cofense / Keepnet |
| Executives targeted vs. average employees | 40–42× more frequently | Recorded Future |
| % of malicious Microsoft 365 docs with QR codes | 83% | Barracuda Networks |
| Unique malicious QR codes in attachments (2025) | 1.7 million | ZenSec |
| Image-based phishing attack increase into 2025 | +400% | APWG |
| Users who scan without verification | 73% | KnowBe4 / NordVPN |
| Brands targeted by QR phishing (Q2 2025) | 1,642 | Mimecast / APWG |
To visualize how rapidly this vector is scaling and where these malicious codes are primarily hidden, see the data distributions below:
Before exploring specific real-world execution methods, it is crucial to understand how these deployment tactics break down across core delivery channels globally:
Understanding the anatomy of a quishing attack is the first step to recognizing one.
Phase 1 – Preparation: The attacker generates a malicious QR code pointing to a credential-harvesting site or malware download, builds a spoofed website replicating a real login page, and often routes the URL through trusted redirect services to mask the destination.
Phase 2 – Delivery: The QR code reaches the target through email (embedded in the body or a PDF attachment), physical stickers on parking meters or restaurant tables, postal mail, or messaging apps.
Phase 3 – The Scan: Because the attack forces the victim to use their personal mobile phone, the entire corporate security perimeter is bypassed. The scan moves from a managed corporate desktop protected by endpoint detection, web proxy, and DNS filtering to a personal smartphone with none of those controls.
Phase 4 – Theft: The victim is prompted to enter Microsoft 365 credentials, banking details, or credit card numbers, or downloads malware masquerading as a legitimate app.
Phase 5 – Post-Compromise: Attackers log into corporate email, launch BEC attacks, sell session tokens on dark web markets, or deploy ransomware across the network.
Scammers print professional-looking QR code stickers and paste them directly over the legitimate codes on parking meters. The stickers are laminated to look official and designed to blend seamlessly into the meter’s design.
What happens: A driver scans what appears to be the official parking code, lands on a spoofed payment site (e.g., “poybyphone[.]com” instead of “paybyphone[.]com”), enters their name, vehicle details, and full payment card number, which goes directly to the attacker.
Real example: In Austin, Texas, scammers placed fake QR code stickers on dozens of city parking meters, directing drivers to a fraudulent payment site to capture credit card details. Similar campaigns hit ParkByPhone meters across multiple UK cities.
Red flag: Look for stickers placed over the original meter surface. The QR code should be integrated into the panel design, not adhered with an overlay sticker.
Attackers physically visit retail environments and overlay legitimate QR codes with their own. Customers at restaurants scan what appears to be the menu QR code and instead land on a fake coupon redemption page requesting payment card “verification.” Some variants trick users into downloading trojanized loyalty apps.
This attack works precisely because users are in a relaxed, trusted environment; entering card details on a QR-linked page feels like a natural extension of the transaction already underway.
Documented case: Fake QR code stickers placed over legitimate codes at 200 store locations caused a 15% drop in legitimate scans and $2.3 million in damage control costs in 2025.
This is the dominant enterprise attack vector. Attackers embed a QR code in a phishing email body or inside a PDF/DOCX attachment. The email’s subject line impersonates HR, IT, payroll, or compliance processes.
Common subject lines observed:
Why it bypasses gateways: The QR code is an image file. The gateway sees a JPEG with no suspicious URLs, passes it as clean, and delivers it to the inbox. The victim scans with their personal phone entirely outside the corporate security perimeter.
Case study (Sophos): A documented campaign used subject lines like “2024 financial plans,” “benefits open enrollment,” and “dividend payout.” The QR codes were embedded inside PDF attachments, an evolution designed to defeat even email gateway image scanning.
Case study (Corporate breach): A South Asian sustainability platform suffered a full Microsoft 365 compromise after staff received a “bonus notification” email with a QR code. Scanning it led to a spoofed login page. After credential theft, the attacker set up invisible email forwarding rules to continuously exfiltrate communications.
Nation-state example: The FBI issued a formal IC3 Flash Alert in January 2026 warning that North Korea’s Kimsuky APT group is using malicious QR codes in spear-phishing campaigns against academics, government employees, think tanks, and defense contractors, redirecting victims to mobile-optimized credential-harvesting pages and collecting device fingerprint data.
Some attacks use the QR code as a direct malware delivery mechanism. The QR code initiates a malicious APK download disguised as a logistics app, delivery tracker, utility payment app, or security update mobile devices.
Kimsuky Android RAT campaign (Dec 2025 – Jan 2026): Kimsuky distributed trojanized Android APKs via QR codes on phishing websites. Sites displayed: “Open this page on your phone scan the QR code.” When users installed the “app,” it granted Kimsuky remote access to camera, microphone, contacts, and SMS messages.
This advanced technique combines quishing with Adversary-in-the-Middle (AitM) attacks to defeat Multi-Factor Authentication and bypass security operations centers entirely:
| Step | Action |
|---|---|
| 1 | Victim scans QR code in phishing email |
| 2 | Redirected to attacker’s reverse proxy server (between victim and real Microsoft/Google) |
| 3 | Victim enters real credentials forwarded live to the real server |
| 4 | Microsoft sends legitimate MFA push to the victim’s phone |
| 5 | Victim approves MFA, believing it’s a real login |
| 6 | Attacker captures the fully-authenticated session token including MFA |
| 7 | Attacker accesses the account without triggering MFA again |
PhaaS kits like Tycoon 2FA and EvilProxy have this capability built-in, lowering the skill bar for attackers dramatically.
| Evasion Technique | Description | Bypass Goal |
|---|---|---|
| QR in PDF attachment | QR code embedded inside a PDF, not the email body | Defeats email body image scanning |
| ASCII-art QR codes | QR rendered using ASCII/Unicode text characters | Defeats image-based detection |
| “Fancy” QR with logos | Brand logos embedded inside custom QR designs | Disrupts pixel-pattern analysis |
| Blob URI redirects | Malicious page rendered from browser-generated data object | Avoids server-hosted URL reputation checks |
| Legitimate redirect chains | Routes through Google, Bitly, Cloudflare before malicious page | Passes URL blocklists |
| Split-image / multipart MIME | QR split across fragments assembled by email client | Defeats single-image scanning |
| Deep link / in-app URLs | Opens inside Telegram; WeChat bypasses mobile browser | Avoids mobile browser URL inspection |
| Dynamic QR codes | Destination URL changes after gateway check | Defeats time-of-scan analysis |
| CAPTCHA-gated landing pages | Protected by Cloudflare Turnstile or reCAPTCHA | Defeats automated sandbox analysis |
| Payload Category | What Is Captured | Typical Impact |
|---|---|---|
| Corporate Credentials | Microsoft 365, Google Workspace, Okta, VPN logins | Account takeover, data breach, ransomware |
| Payment Card Details | Card number, CVV, expiry, billing address | Financial fraud, card cloning |
| Banking Credentials | Online banking username, password, PIN | Direct fund transfers |
| Session Tokens / Cookies | Authenticated session after MFA | MFA bypass, persistent unauthorized access |
| PII | Name, SSN/Aadhaar, address, date of birth | Identity theft, credit fraud |
| UPI / Mobile Payment Data | UPI PIN, transaction confirmation codes | Unauthorized UPI transfers (India) |
| Mobile Device Access | Camera, microphone, contacts, SMS (via APK) | Espionage, OTP interception |
| Corporate Secrets | Email contents, documents | IP theft, BEC fraud |
| Warning Sign | What to Look For |
|---|---|
| Sticker overlay | QR on a sticker placed on top of the original surface |
| Misaligned placement | Code positioned in an unusual spot |
| Inconsistent branding | Different font, colors, or logo than surrounding signage |
| Brand-new code on worn signage | Fresh pristine sticker on aged signage |
| Suspicious URL preview | Domain with typos, hyphens, random strings |
| Requests for unnecessary data | Asking for SSN, full card + CVV together |
| Urgent call-to-action | “Scan now or lose access” language |
| Warning Sign | What to Look For |
|---|---|
| QR instead of a link | Legitimate companies rarely replace links with QR codes |
| External sender domain | Doesn’t match the company being impersonated |
| Urgency language | “Verify within 24 hours,” “Account will be suspended” |
| QR inside a PDF attachment | Major red flag legitimate services don’t do this |
| Unsolicited HR/IT/payroll theme | Benefits enrollment or MFA reset from unexpected senders |
| Generic salutation | “Dear User” instead of your actual name |
| Industry | Risk Level | Primary Attack Goal |
|---|---|---|
| Energy | Critical (29% of malware QR emails) | Operational disruption, credential theft |
| Financial Services | Critical | Banking credentials, payment fraud |
| Healthcare | Critical | Patient data, ransomware |
| Technology | High | Intellectual property, SaaS access |
| Manufacturing | High | OT/ICS access, ransomware |
| Education | High (15,000+ daily phishing QR emails) | Staff/student credential theft |
| Retail | High (highest employee miss rate) | Customer payment data |
| Government / Defense | High | Espionage, classified data |
| Country | Reporting Channel |
|---|---|
| United States | FBI IC3 at ic3.gov; FTC at reportfraud.ftc.gov |
| United Kingdom | Action Fraud at actionfraud.police.uk |
| India | cybercrime.gov.in or call 1930 |
| EU | ENISA or national CERT; local police |
| Australia | cyber.gov.au / ReportCyber portal |
| Control | Description |
|---|---|
| AI-powered email security with OCR | Deploy platforms with image recognition that decode QR codes in emails AND PDF attachments |
| Phishing-resistant MFA | Enforce FIDO2/hardware keys defeats AitM session token theft |
| Conditional Access policies | Block unmanaged devices from accessing sensitive corporate resources |
| MDM / Mobile Device Management | Restrict QR scanning to approved apps on enrolled corporate devices |
| Quishing-specific security training | Run QR phishing simulations improves detection by ~87% in 3 months |
| DMARC / DKIM / SPF | Harden email authentication to block sender spoofing |
| Short session lifetimes | Sessions expire quickly; use continuous session validation and device-bound tokens |
| Zero Trust architecture | Validate sessions continuously based on device, location, and behavior |
| Physical QR code audits | Train staff to regularly inspect and test all QR codes on premises |
| Incident response playbook | Include QR-specific steps: session revocation, credential reset, device isolation |
Google has released an important Chrome 153 security update that fixes 42 vulnerabilities across the…
The Cybersecurity and Infrastructure Security Agency (CISA) and the National Institute of Standards and Technology…
CISA and five international cybersecurity agencies have released detailed guidance describing 17 common techniques hackers…
Apple has released one of its largest coordinated security rollouts, addressing 273 distinct critical vulnerabilities…
You can’t detect today's attacks with yesterday’s threat intelligence; that’s how you could briefly formulate…
Microsoft has published a draft Humanist AI Code of Conduct that would prohibit its in-house…