Homograph attack
A homograph attack uses visually confusable characters to make one identifier, such as a domain name, look like another and mislead the person reading it.
Similar appearance does not mean the same name
Different characters can look alike within the same alphabet or across writing systems. An attacker can exploit that resemblance in a domain name. The address may look familiar without being the address the person intended to visit.
This differs from ordinary typosquatting, which anticipates a typing or spelling mistake. A homograph can fool someone reading a link even when they have not typed anything.
Internationalized domains are legitimate
Internationalized domain names allow people to use more than the basic Latin alphabet. They are not inherently suspicious. ICANN’s guidance recognizes both spoofing concerns and the need to preserve legitimate use of these names.
Browsers apply display rules to reduce confusion. Chromium documents when Chrome displays a name in Unicode and when it instead shows ASCII-compatible labels beginning with xn--. Those labels use Punycode encoding. Punycode is not encryption or, by itself, evidence of a malicious site.
Do not make visual inspection your only check before a sensitive action. Use an established bookmark for a known service and confirm unexpected document requests through a separate trusted route. The shared-document article explains why a familiar-looking sharing message should not decide where you sign in.
For responders, retain the actual URL rather than relying only on a screenshot. The underlying characters matter when distinguishing a deceptive domain from the legitimate one.