Direct answer
Invisible Unicode characters are real code points that may be present in a string without appearing as ordinary visible letters or symbols. Examples include ZERO WIDTH SPACE, WORD JOINER, directionality controls, and other formatting characters.
Why humans and machines disagree
A human typically sees glyphs painted by a renderer. Software can instead inspect code points, UTF-8 bytes, normalization results, token boundaries, or control characters. Two strings that look the same can therefore be different at the machine level.
Detection
A code-point inspector can reveal the exact character sequence. Security-oriented analysis often compares the raw string with normalized views and flags format controls, unusual whitespace, mixed scripts, or bidirectional controls rather than blindly deleting them.
Puzzle relevance
Invisible characters make good educational carriers because the hidden state can be almost impossible to notice visually while remaining exact and deterministic for software.
Frequently asked questions
Does normalization remove invisible Unicode?
Not necessarily. Unicode normalization and security sanitization are different operations, and many format controls survive NFC and NFKC.
Are invisible characters always malicious?
No. Many have legitimate typographic, linguistic, and line-breaking uses.