#endpoint-protection — Public Fediverse posts

The Death of Trust: Why Adversary-in-the-Middle and QR Attacks Are Gutting Corporate Defenses

1,408 words, 7 minutes read time.

We have built our enterprise security perimeters on a foundation of sand, and the storm has finally arrived. For years, security leadership assured stakeholders that multi-factor authentication was the ultimate shield against unauthorized entry. That comfort was a lie, born of complacency and soft thinking. Threat actors have evolved past simple credential harvesting; they are actively deploying Adversary-in-the-Middle (AiTM) phishing frameworks and zero-visibility QR code vectors that bypass legacy defenses without breaking a single line of code. We watched the systems fail because we relied on passive compliance rather than uncompromising operational discipline.

The modern attack landscape does not care about our corporate policy manuals or annual security awareness videos. According to data tracked by Microsoft Threat Intelligence and the Cybersecurity and Infrastructure Security Agency (CISA), targeted organizations face an unprecedented surge in AiTM token hijacking and image-embedded QR phishing campaigns designed specifically to bypass secure email gateways. Threat actors set up reverse-proxy infrastructure that sits directly between our users and legitimate authentication endpoints. When our workforce logs in, the attacker proxies the request, steals the authenticated session cookie in real time, and walks straight through the front door.

We must strip away the illusion that traditional defensive controls are keeping us safe. The rise of hybrid voice-and-email callback attacks and CAPTCHA-gated phishing infrastructure proves that attackers are exploiting the exact gaps we left open through institutional laziness. This deep-dive analysis confronts the mechanical realities of modern social engineering, dissects the failure of static multi-factor authentication, and lays out the hard architecture required to survive.

The Mechanics of Session Hijacking: How Reverse Proxies Bypass Legacy MFA

We need to understand the structural failure occurring at the authentication layer. Legacy multi-factor authentication relies on push notifications, SMS codes, or time-based one-time passwords (TOTP) to verify identity. Threat actors bypass these controls entirely by utilizing advanced AiTM phishing proxies like EvilGinx. The attack vector does not attempt to break encryption algorithms or crack complex passwords; it simply manipulates the human operating the endpoint into handing over full access.

The process is surgically precise and brutally efficient. An employee receives a high-urgency communication, often mimicking internal human resources or critical SaaS infrastructure. The link inside routes the target to an attacker-controlled proxy server that dynamically mirrors the organization’s real login portal in real time. As the employee inputs their primary credentials and completes the secondary MFA challenge, the proxy server relays those packets directly to the legitimate service provider.

Once the service provider authenticates the session, it issues a session token or authentication cookie back to the user. Because that traffic flows directly through the adversary’s proxy, the attacker captures the live session token instantaneously. The attacker then injects that session token into their own browser, instantly hijacking the active, fully authenticated corporate session without ever needing to know the user’s password or possess their physical MFA token. We must realize that the authentication process completed successfully in the eyes of the server, yet the enterprise was compromised entirely.

Blind Spots in the Perimeter: The Rise of Image-Based Quishing and TOAD Attacks

We have allowed our perimeter defenses to become completely blind to non-traditional attack vectors. Standard secure email gateways spend massive compute resources scanning incoming plain-text URLs and parsing standard attachments for known malicious signatures. Cybercriminals recognize this dependency and have adapted by shifting their payloads into optical and multi-channel vectors that bypass automated text-scanning engines entirely.

QR code phishing, known colloquially as quishing, represents a massive operational vulnerability across corporate environments. Attackers embed malicious authentication URLs inside image-based QR codes wrapped within benign PDF documents or inline email bodies. Traditional email filters read the image as harmless static graphic data, allowing the message to land directly in the victim’s inbox. The target is instructed to scan the code using a personal mobile device to complete an mandatory system update or open a secure document.

This action immediately forces the target off our managed corporate network and onto an unmonitored mobile device that lacks endpoint detection and response (EDR) agents. From there, the mobile browser connects to the AiTM proxy framework, and the session is compromised out of sight of internal security operations. Simultaneously, Telephone-Oriented Attack Delivery (TOAD) campaigns combine fake subscription invoices with live callback centers. When employees call the listed support number, live operators leverage high-urgency social engineering to guide targets through manual credential entry or remote control software installation. The system fails because we trained our people to look for suspicious links while leaving them exposed to visual and verbal exploitation.

Engineering Absolute Resistance: Deploying FIDO2 and Zero-Trust Architecture

We cannot solve an architectural flaw with soft solutions like additional compliance training or stern warning banners. Human error under psychological manipulation is an inevitable reality; therefore, our technical controls must enforce absolute resistance at the protocol level. We must eliminate push-based and TOTP-based authentication schemes in favor of hardware-bound, phishing-resistant credentials based on FIDO2 and WebAuthn standards.

FIDO2 architecture neutralizes AiTM reverse-proxy attacks through cryptographic origin binding. When an employee authenticates using a FIDO2 security key or platform authenticator, the web browser cryptographically binds the authentication request to the specific domain origin shown in the browser address bar. If an attacker routes the employee to a spoofed proxy domain, the browser detects the domain mismatch and refuses to sign the authentication challenge. The proxy captured nothing because no valid token was ever generated.

Beyond hardware-bound authentication, we must enforce strict Zero-Trust network access and continuous conditional access policies. Authentication must never be treated as a single, static event that grants permanent access for the duration of a session token. System infrastructure must continuously evaluate device health, geographic velocity, IP reputation, and behavioral anomalies throughout the entire life of the connection. If a session token suddenly presents from an unrecognized IP address or an unmanaged device, that session must be revoked immediately, and step-up authentication must be enforced.

Summary of Core Defenses

• Hardware-bound FIDO2 and WebAuthn protocols eliminate AiTM session hijacking by cryptographically binding authentication challenges to verified domain origins.
• Optical payload inspection and mobile device management controls counter image-based quishing vectors that bypass traditional secure email gateways.
• Continuous conditional access policies validate device compliance, network context, and session integrity in real time to neutralize stolen session tokens.

The era of trusting simple passwords paired with basic multi-factor push notifications is dead. Attackers have weaponized automated proxies and optical vectors to turn our own authentication mechanisms against us. Survival requires dropping the comfortable illusion of legacy security, taking absolute accountability for our technical debt, and building an unyielding architecture that stops threat actors cold before the system burns.

D. Bryan King

Sources

• CISA: Defending Against Software Supply Chain Attacks & Phishing Proxies
• Microsoft Security: From Cookie Theft to BEC: Adversary-in-the-Middle Phishing
• MITRE ATT&CK: Steal Web Session Cookie (Technique T1539)
• NIST National Vulnerability Database: Session Hijacking & Token Exploits
• FIDO Alliance: FIDO2 Authentication Standards & Specifications
• W3C: Web Authentication (WebAuthn) Specification
• FBI Internet Crime Complaint Center (IC3): Public Service Announcement on Quishing Trends
• SANS Institute: Combatting Adversary-in-the-Middle (AiTM) Phishing Attacks
• NIST Special Publication 800-63-3: Digital Identity Guidelines
• Mandiant: Dissecting Adversary-in-the-Middle Infrastructure
• BleepingComputer: Microsoft Warns of Massive AiTM Phishing Campaigns
• Krebs on Security: Microsoft on AiTM Phishing Attacks That Bypass MFA
• Dark Reading: Quishing on the Rise: How QR Code Attacks Bypass Defenses
• SecurityWeek: QR Code Phishing Trends Targeting Corporate Enterprise Networks
• Infosecurity Magazine: MFA Bypass and Session Token Hijacking Surge
• CrowdStrike: What is an Adversary-in-the-Middle (AiTM) Attack?
• Palo Alto Networks: Understanding AiTM Phishing Frameworks and Reverse Proxies
• Trend Micro: Adversary-in-the-Middle Phishing Attacks Explained
• Fortinet: What is Quishing? Understanding QR Code Phishing Threats
• Proofpoint: Telephone-Oriented Attack Delivery (TOAD) Threat Insight
• Rapid7: Understanding Session Hijacking and Pass-the-Cookie Attacks
• Check Point Software: Comprehensive Guide to Quishing and EvilGinx Exploits
• Sophos: Technical Report on Modern MFA Bypass Techniques
• SentinelOne: Session Cookie Theft: How Attackers Bypass MFA
• ThreatDown: QR Code Phishing (Quishing) Mechanics Explained
• Schneier on Security: Analysis of MFA Bypasses via Reverse Proxies
• ZDNET: QR Code Phishing as a Growing Enterprise Security Threat
• The Hacker News: Analysis of EvilGinx3 Phishing Framework Capabilities
• CNET: Quishing: How Scammers Use QR Codes to Steal Credentials
• Aruba Networks: Fundamentals of Zero Trust Network Access (ZTNA) Architecture

Disclaimer:

The views and opinions expressed in this post are solely those of the author. The information provided is based on personal research, experience, and understanding of the subject matter at the time of writing. Readers should consult relevant experts or authorities for specific guidance related to their unique situations.

Related Posts

Rate this:

ICYMI: A security researcher recently published a working tool that extracts credentials stored in Microsoft Edge directly from the browser's parent process memory. No exploit is needed – just sufficient system privileges.

This is the kind of threat Keeper Security is designed to help address. In addition to our secure and encrypted password manager, Keeper Forcefield blocks unauthorized memory access at the kernel level – so even if a machine is compromised, there's nothing to extract.

#KeeperSecurity #Cybersecurity #PasswordSecurity #EndpointProtection #MicrosoftEdge

#checkpoint basically says whoever is developing #public #internet facing software needs #AI to find #exploits before others find them (any program that exchanges messages)
#firewall #virusscanner #endpointprotection good but not enough

https://blog.checkpoint.com/artificial-intelligence/claude-mythos-wake-up-call-what-ai-vulnerability-discovery-means-for-cyber-defense/ #cyber #cybersecurity

In 2025, the top 5 known exploited vulnerability (#KEV) vendors as of https://cyble.com/blog/cisa-kev-2025-exploited-vulnerabilities-growth/ were:

Microsoft (39)
#Apple (9)
#Cisco (8)
#Fortinet (8)
#Google #Chromium (7)

If you like to minimize your #security risk here, avoiding those vendors could improve your overall exposure.

As you can see, this is particularly true for #Microsoft.

Mitigation using #AntiMalware or #EndPointProtection is not the answer as we've learned in the previous year where the "Most Frequently Exploited #Vulnerabilities" have been security products!
Source: https://services.google.com/fh/files/misc/m-trends-2025-en.pdf

If you have high requirements for #ITsecurity, you need to migrate your systems to #Linux which is also part of KEV but on a *much* better level!

#Windows #macOS #iOS #exploits

WhatsApp accounts targeted in ‘GhostPairing’ attack https://www.csoonline.com/article/4108925/whatsapp-accounts-targeted-in-ghostpairing-attack.html #CommunicationsSecurity #EndpointProtection #MessagingSecurity #NetworkSecurity #MobileSecurity #Security

Hundreds of Ivanti EPM systems exposed online as critical flaw patched https://www.csoonline.com/article/4104158/hundreds-of-ivanti-epm-systems-exposed-online-as-critical-flaw-patched.html #EndpointProtection #Vulnerabilities #Security

Gemini for Chrome gets a second AI agent to watch over it https://www.csoonline.com/article/4103346/gemini-for-chrome-gets-a-second-ai-agent-to-watch-over-it-2.html #ArtificialIntelligence #EndpointProtection #BrowserSecurity #Security

Neue bösartige Browser-Erweiterungen entdeckt https://www.csoonline.com/article/4100071/neue-bosartige-browser-erweiterungen-entdeckt.html #EndpointProtection #BrowserSecurity #Security

Newly discovered malicious extensions could be lurking in enterprise browsers https://www.csoonline.com/article/4099446/newly-discovered-malicious-extensions-could-be-lurking-in-enterprise-browsers.html #EndpointProtection #BrowserSecurity #Cybercrime #Security #Malware

Microsoft gives Windows admins a legacy migration headache with WINS sunset https://www.csoonline.com/article/4098729/microsoft-gives-windows-admins-a-legacy-migration-headache-with-wins-sunset-2.html #NetworkAdministrator #EndpointProtection #NetworkSecurity #WindowsSecurity #Security #Careers #Roles

AI browsers can be tricked with malicious prompts hidden in URL fragments https://www.csoonline.com/article/4097087/ai-browsers-can-be-tricked-with-malicious-prompts-hidden-in-url-fragments.html #ArtificialIntelligence #EndpointProtection #BrowserSecurity #Security

New ClickFix attacks use fake Windows Update screens to fool employees https://www.csoonline.com/article/4096241/new-clickfix-attacks-use-fake-windows-update-screens-to-fool-employees.html #EndpointProtection #WindowsSecurity #Cybercrime #Security #Malware

Recognizing and responding to cyber threats: What differentiates NDR, EDR and XDR https://www.csoonline.com/article/4091641/recognizing-and-responding-to-cyber-threats-what-differentiates-ndr-edr-and-xdr.html #SecurityInfrastructure #EndpointProtection #RiskManagement #Security

Sneaky2FA phishing tool adds ability to insert legit-looking URLs https://www.csoonline.com/article/4094165/sneaky2fa-phishing-tool-adds-ability-to-insert-legit-looking-urls.html #EndpointProtection #SocialEngineering #BrowserSecurity #Cybercrime #Phishing #Security

Hidden API in Comet AI browser raises security red flags for enterprises https://www.csoonline.com/article/4092995/hidden-api-in-comet-ai-browser-raises-security-red-flags-for-enterprises.html #EndpointProtection #BrowserSecurity #Security

More work for admins as Google patches latest zero-day Chrome vulnerability https://www.csoonline.com/article/4092287/more-work-for-admins-as-google-patches-latest-zero-day-chrome-vulnerability.html #Zero-DayVulnerabilities #EndpointProtection #BrowserSecurity #Vulnerabilities #Security

🚨 New sophisticated threat alert! RONINGLOADER leverages signed drivers to disable Microsoft Defender & bypass EDR, targeting Chinese users with multi-stage malware. Stay informed & protected! 🛡️🔒 #CyberSecurity #Malware #InfoSec #EndpointProtection https://gbhackers.com/roningloader/
#newz

Microsoft fixes Windows 10 update flaw https://www.csoonline.com/article/4091597/microsoft-fixes-windows-10-update-flaw.html #EndpointProtection #WindowsSecurity #Security

Russian APT abuses Windows Hyper-V for persistence and malware execution https://www.csoonline.com/article/4085272/russian-apt-abuses-windows-hyper-v-for-persistence-and-malware-execution.html #AdvancedPersistentThreats #EndpointProtection #WindowsSecurity #Cyberattacks #Cybercrime #Security

Chinese hackers target Western diplomats using hard-to-patch Windows shortcut flaw https://www.csoonline.com/article/4082701/chinese-hackers-target-western-diplomats-using-hard-to-patch-windows-shortcut-flaw.html #EndpointProtection #Vulnerabilities #WindowsSecurity #Security

Critical Microsoft WSUS flaw exploited in wild after insufficient patch https://www.csoonline.com/article/4078952/critical-microsoft-wsus-flaw-exploited-in-wild-after-insufficient-patch.html #EndpointProtection #Vulnerabilities #WindowsSecurity #Security

AI browsers can be abused by malicious AI sidebar extensions: Report https://www.csoonline.com/article/4078306/ai-browsers-can-be-abused-by-malicious-ai-sidebar-extensions-report-2.html #EndpointProtection #BrowserSecurity #Cybercrime #Security #Malware

Security patch or self-inflicted DDoS? Microsoft update knocks out key enterprise functions https://www.csoonline.com/article/4076016/security-patch-or-self-inflicted-ddos-microsoft-update-knocks-out-key-enterprise-functions-2.html #DataandInformationSecurity #EndpointProtection #WindowsSecurity #Encryption #Security

Network security devices endanger orgs with ’90s era flaws https://www.csoonline.com/article/4074945/network-security-devices-endanger-orgs-with-90s-era-flaws.html #DataandInformationSecurity #ApplicationSecurity #EndpointProtection #TechnologyIndustry #NetworkSecurity #Vulnerabilities #DevSecOps #Industry #Security #Markets

OT security: Why it pays to look at open source – Source: www.csoonline.com https://ciso2ciso.com/ot-security-why-it-pays-to-look-at-open-source-source-www-csoonline-com/ #rssfeedpostgeneratorecho #EndpointProtection #CyberSecurityNews #IncidentResponse #NetworkSecurity #CSOonline #CSOOnline

OT security: Why it pays to look at open source https://www.csoonline.com/article/4055176/ot-security-why-it-pays-to-look-at-open-source.html #EndpointProtection #IncidentResponse #NetworkSecurity

10 most powerful cybersecurity companies today – Source: www.csoonline.com https://ciso2ciso.com/10-most-powerful-cybersecurity-companies-today-source-www-csoonline-com/ #IdentityManagementSolutions #IntrusionDetectionSoftware #rssfeedpostgeneratorecho #EndpointProtection #CyberSecurityNews #NetworkSecurity #riskmanagement #accesscontrol #CSOonline #Microsoft #zerotrust #CSOOnline #Security

10 most powerful cybersecurity companies today https://www.csoonline.com/article/569075/the-10-most-powerful-cybersecurity-companies.html #IdentityManagementSolutions #IntrusionDetectionSoftware #EndpointProtection #NetworkSecurity #RiskManagement #AccessControl #Microsoft #ZeroTrust #Security

CrowdStrike: A new era of cyberthreats from sophisticated threat actors is here – Source: www.csoonline.com https://ciso2ciso.com/crowdstrike-a-new-era-of-cyberthreats-from-sophisticated-threat-actors-is-here-source-www-csoonline-com/ #ThreatandVulnerabilityManagement #rssfeedpostgeneratorecho #EndpointProtection #CyberSecurityNews #CloudSecurity #CSOonline #CSOOnline