A web address sometimes appears in readable international characters in one place and as a label beginning with xn-- in another. That change can look like corruption, but it can be a normal part of how internationalised domain names are represented.
ICANN calls the ASCII-compatible form an A-label. Punycode is the encoding used within this system to represent international characters using a restricted set of ASCII characters. The readable Unicode form is commonly called a U-label.
One name, different representations
For an illustrative reserved example, bücher.example can be represented as xn--bcher-kva.example. The .example ending here is for explanation, not a recommendation to visit a real service. The encoded part represents the label before the dot; it is not a password or a translated description of the website.
This supports domain names in languages that need characters beyond the basic Latin letters. ICANN’s introduction to internationalised domain names provides that wider context.
Encoding does not establish trust
Punycode is not encryption, and seeing xn-- does not by itself show that a site is malicious. Equally, a familiar-looking Unicode name is not proof that a site belongs to the organisation you expect. The representation and the identity are separate questions.
When checking an important address, compare the actual hostname with a trusted record or bookmark. Do not rely only on the words printed in an email link. If you are uncertain, open the organisation through a route you already trust rather than deciding from the visual appearance of one unfamiliar label.
Different applications can choose different display forms, so a copied address may look different after pasting into another tool. For a support request, provide the complete address and describe where you saw each form. That gives the person helping you something precise to compare without assuming every international character indicates a problem.