| Unvocalised text, ambiguous reading | With vowels unwritten, one consonant string can be read several ways; a wrong reading of a name or term changes the document’s meaning. | A native specialist resolves the reading from context and the source document, not by guesswork — and confirms names against the ID. |
| Transliteration that must match the passport | Several romanisation systems exist; a Hebrew name can become כהן → Cohen / Kohen / Kohn. Only the passport spelling is accepted. | Names are transliterated to match the passport exactly, not respelled by a generic scheme. |
| Bidirectional (RTL/LTR) breakage | Mixed Hebrew and Latin, numbers, dates and codes can reverse or reorder when handled by non-RTL tools — a real error in certificates and contracts. | Layout is set with correct bidirectional handling so Hebrew, Latin, numbers and codes all read in the right order. |
| Final letter forms | Five letters change shape word-finally (ך ם ן ף ץ). Using the wrong form marks the text as non-native and can distort a name. | Final and medial forms are set correctly throughout, so the text reads as native Hebrew. |
| Modern vs Biblical/liturgical register | Modern Hebrew differs from Biblical and liturgical Hebrew; the wrong register in an official document reads as archaic or off-key. | The correct Modern Hebrew register is used for official and business documents, distinct from liturgical language. |
| Hebrew dates & the calendar | Israeli documents may carry Hebrew-calendar dates; a mechanical conversion misstates the day. | Dates are converted carefully between the Hebrew and Gregorian calendars and cross-checked against the source. |