In this article
- What EvilProxy is — and what it is not
- How the attack works
- The targets
- An organized and accessible criminal service
- What this means for protecting your accounts
- A citizen perspective: understanding without overestimating the risk
- Caveats and limits of this article
- Frequently asked questions
- Further reading
Two-factor authentication is often presented as a solid barrier against account hijacking. Yet a tool that emerged in cybercriminal circles, known as EvilProxy, shows that no security measure is absolute. Understanding how it works gives you a clearer view of how well your online accounts are really protected.
What EvilProxy is — and what it is not
EvilProxy is a phishing-as-a-service (PhaaS) platform. It allows malicious actors, even those with little technical experience, to run sophisticated phishing campaigns capable of bypassing two-factor authentication (2FA). It is not a virus or malware installed on your device, but a criminal infrastructure rented on demand.
This type of tool was documented by researchers at the firm Resecurity, who published a detailed analysis of how it works. In cybercriminal circles, EvilProxy was marketed as an accessible way to compromise accounts protected by 2FA — a security measure many users wrongly consider airtight.
How the attack works
EvilProxy relies on a technique known as a malicious reverse proxy, sometimes called an Adversary-in-the-Middle (AiTM) attack. Here is the principle:
- The victim receives a phishing link leading to a cloned page of a legitimate service (Google, Microsoft, Facebook, Apple iCloud, Dropbox, GitHub and many others).
- When they enter their credentials and 2FA code, the data passes in real time through a proxy server controlled by the attacker.
- The server forwards the request to the real service, receives the authenticated session and captures it.
- The attacker obtains a valid session cookie — which gives access to the account without needing the 2FA code, which has already been used.
The trick is that the victim really does sign in to the real service. They never deal with a simple static copy: everything looks normal, except that their session is intercepted along the way.
This method differs from classic phishing. A 2FA code entered on a static fake page would be useless to the attacker, because the code expires within seconds. Here, the proxy acts in real time and uses the code at the very moment the victim submits it.
The targets
The list of services targeted by EvilProxy is particularly long. Published analyses mention, among others:
- Google and Gmail
- Microsoft (personal and work accounts)
- Apple iCloud
- Facebook and Instagram
- Dropbox
- GoDaddy
- GitHub, NPM, PyPI and RubyGems
- Yahoo and Yandex
The presence of software development platforms — GitHub, NPM, PyPI, RubyGems — deserves special attention. These targets suggest that some operators go after not only individuals but also developers and engineers, in order to compromise widely used software projects. Unauthorized access to these accounts could allow malicious code to be injected into shared libraries or projects, with cascading effects on everyone who uses that software — this is known as a software supply chain attack.
Software supply chain attacks are especially feared because they affect not only the direct victim but potentially everyone who uses the compromised software or libraries.
An organized and accessible criminal service
What makes EvilProxy worrying is not only its technical sophistication but its commercial structure. It is a subscription service, available to malicious actors without advanced technical expertise.
Documented pricing model
Analyses published at the time of discovery reported subscriptions of 10, 20 or 31 days, with prices varying by targeted service — and higher prices for Google accounts. These amounts reflected the situation when the research was first published and may have changed.
Access and distribution
The service was distributed through the TOR anonymity network, with payments arranged over encrypted messaging. Access required manual approval, in theory limiting customers to actors known in those circles.
This business model reflects a broader trend: the professionalization of cybercrime tools. Skills that were once reserved for highly qualified attackers are now available to anyone with the budget and the right contacts. The barrier to entry is falling, which mechanically increases the volume and variety of actors able to launch this kind of attack.
What this means for protecting your accounts
The existence of tools like EvilProxy does not mean 2FA is useless. It remains a significant layer of protection against a large share of common attacks. But it invites you to understand its limits and adopt complementary practices.
2FA considerably reduces the risk of intrusion in most scenarios. It is not, however, absolute protection against real-time attacks like those EvilProxy makes possible. Caution with the links you receive remains the first line of defence.
Practices that still matter
- Never click a link received by email or message to access a sensitive account. Always type the address directly into the browser or use a saved bookmark.
- Check the URL in the address bar before entering anything. A malicious proxy can imitate an interface very closely, but the domain address will be different.
- Use physical security keys or passkeys (FIDO2/WebAuthn) where possible. These methods are bound to the original domain and cannot be intercepted by a proxy, unlike TOTP codes.
- Turn on sign-in alerts for services that offer them, so you are told quickly about unusual access.
- Use a password manager: it will not offer to autofill your credentials on a domain it does not recognize, which acts as an implicit warning.
What is less effective against this kind of attack
- One-time codes (SMS or TOTP app) can be intercepted in real time by an active proxy.
- The HTTPS padlock alone does not guarantee a site is legitimate — phishing pages also use valid certificates.
A citizen perspective: understanding without overestimating the risk
Tools like EvilProxy primarily target high-value accounts: work accounts, access to company systems, developers with rights to shared code repositories. Ordinary individuals are not out of reach, but they are generally less exposed than targets with high strategic or financial value.
That said, the spread of these tools is changing the landscape. Phishing campaigns once reserved for specific targets can now be deployed on a larger scale. Digital vigilance is no longer just a matter for IT security professionals.
From a household preparedness perspective, digital security is part of everyday resilience, just like physical or food security. Understanding common attack vectors helps you make informed decisions without slipping into anxiety or paranoia.
The goal is not to become a cybersecurity expert, but to adopt a few lasting habits that significantly reduce your exposure to the most frequent attacks — and to recognize the warning signs before you click.
Caveats and limits of this article
EvilProxy was publicly documented and analysed in the fall of 2022. The cybercrime toolkit changes quickly. The technical and pricing details mentioned here reflect the situation at the time of the initial discovery. Variants or successors have probably appeared since. The protection principles, however, remain relevant over time.
- Information about EvilProxy’s prices and access terms comes from research published in 2022 and may be out of date.
- Other similar services have likely emerged since, with comparable or more advanced features.
- The protection recommendations here are based on general cybersecurity principles, not on real-time threat monitoring.
Frequently asked questions
Is 2FA still useful if EvilProxy can bypass it?
Yes. 2FA remains effective against the vast majority of common attacks: sign-in attempts with stolen passwords, credential stuffing, access from an unknown device. EvilProxy is a targeted, active attack that requires the victim to click a malicious link. Caution with the links you receive remains the essential complement to 2FA.
What is the difference between an SMS code, a TOTP app and a physical key?
SMS codes and TOTP apps (such as Google Authenticator or Authy) generate one-time codes valid for a few seconds. A real-time proxy can intercept and use them immediately. FIDO2/WebAuthn physical keys (such as a YubiKey) work differently: they cryptographically sign a request tied to the site’s exact domain. If the domain is different — even by a single character — authentication fails, which makes this kind of attack ineffective against them.
How can I recognize a sophisticated phishing link?
Modern phishing pages can look very convincing. The most reliable clues remain: the domain in the browser’s address bar (look for subtle variations such as “g00gle.com” or an unusual domain), your password manager not offering to autofill, and how you arrived at the page (an unexpected link, artificial urgency, an unusual request).
What should I do if I think I clicked a phishing link?
Act quickly: change the password of the affected account from a different device and connection, revoke active sessions if the service allows it, review recent account activity, and warn your contacts if the compromised account could be used to reach them. If a work account is involved, notify your organization’s IT security lead immediately.
Further reading
Digital security is part of everyday discretion: see our guide to discretion and privacy and our article on “anonymous” data and re-identification.

