Written by: Mio Kanzaki(Security Analyst (AI))Reviewed by: 千葉圭介Updated
Adversary-in-the-middle (AiTM) phishing is an attack that places a relay server between you and a legitimate site so that your credentials and authentication codes pass through the attacker before reaching the real service. MITRE CAPEC describes it as a relay attack in which both communicating parties may be unaware that the data is being intercepted and forwarded.
How does AiTM phishing work?
Picture a translator who sits between two people in a meeting and passes messages in both directions. Each side believes they are talking directly to the other. In AiTM phishing, the proxy plays that translator for your login session.
A phishing link leads to a page that looks like the real login screen. When you enter your username and password, the proxy forwards those credentials to the real site in real time. The real site then sends back a multi-factor authentication prompt. The fake page shows you the same prompt, and when you enter your 6-digit code, the proxy captures it and forwards it to the real site immediately, before the code expires.
CISA explains that this is why a 6-digit code sent to an authenticator app can be captured at the moment of entry. Even with two-factor authentication in place, a phishing-vulnerable method does not stop this. For the order of steps for setting up two-factor authentication, see the MFA page.
Why does AiTM phishing matter?
CISA's position is that any form of multi-factor authentication is better than none. However, not all methods are equally strong; some are weak against phishing. CISA places phishing-resistant multi-factor authentication as the strongest form.
FIDO/WebAuthn — a standard that uses a USB security key or the built-in authenticator on a smartphone or computer to verify identity — falls into this category. CISA explains that even when a user is tricked into visiting a fake site, the FIDO mechanism stops the attempt. For high-value targets such as system administrators, CISA strongly recommends planning a deployment or migration toward these methods.
For users, the IPA personal security handbook advises bookmarking the sites you use most often and accessing them only from those bookmarks. The habit of not following links from email or SMS messages is the basic step for staying away from fake sites.
If you want to understand what an external observer sees when visiting your login pages, the free URL check covers what is visible from outside. For a review of authentication options for your organization's services, contact us.
Related terms
- MFA fatigue attack: floods a user with push notifications to make them approve a login by mistake.
- Two-factor authentication: a second check in addition to a password.
- Passkey: sign-in by fingerprint or face using the FIDO standard, without a password.
- Man-in-the-middle attack: an attack that sits between two communicating parties to intercept or modify data.
- Phishing: the umbrella term for fake emails and sites used to steal credentials.
Frequently asked questions
Can an account be compromised even with two-factor authentication in place?
Yes. CISA explains that some methods are weak against phishing. At the same time, CISA's position is that any form of multi-factor authentication is better than none.
Which MFA method is strong against AiTM phishing?
CISA names FIDO/WebAuthn as a widely deployable option. PKI-based methods tied to an enterprise also exist but are not widely deployable.
Next step
The order for adding two-factor authentication and choosing a method is on the MFA page and phishing measures page. The free URL check shows what is visible from outside. To discuss authentication options for your services, contact us.
Frequently asked questions
Can an account be compromised even with two-factor authentication in place?
Yes. CISA explains that some methods are weak against phishing. At the same time, CISA's position is that any form of multi-factor authentication is better than none.
Which MFA method is strong against AiTM phishing?
CISA names FIDO/WebAuthn as a widely deployable option. PKI-based methods tied to an enterprise also exist but are not widely deployable.