Out of character: Homograph attacks explained

In April, Xudong Zheng, a security enthusiast based in New York, found a flaw in some modern browsers in the way they handle domain names. While Chrome, Firefox, and Opera already have security measures in place to cue users that they might be visiting a destination they thought was legitimate, at that time these browsers did not flag a fake domain name that used all Latin look-alike characters taken from another foreign language. Zheng demonstrated this when he created and registered a proof-of-concept (PoC) page for the domain, аррӏе.com, which was written in pure Cyrillic characters.

What is a homograph attack?

A homograph attack is a method of deception wherein a threat actor leverages on the similarities of character scripts to create and register phony domains of existing ones to fool users and lure them into visiting. This attack has some known aliases: homoglyph attack, script spoofing, and homograph domain name spoofing. Characters—i.e., letters and numbers—that look alike are called homoglyphs or homographs, thus the name of the attack. Examples of such are the Latin small letter O (U+006F) and the Digit zero (U+0030). Hypothetically, one might register bl00mberg.com or g00gle.com and get away with it. But in this day and age, such simple character swaps could be easily detected.

In an internationalized domain name (IDN) homograph attack, a threat actor creates and registers one or several fake domains using at least one look-alike character from a different language. Again, hypothetically, one might register gοοgle.com, but not before swapping the Latin small letter O (U+006F) with the Greek small letter Omicron (U+03BF).

Source: Malwarebytes Labs