Whether blacked-out text can be revealed comes down to one thing: was the original deleted, or just covered up? Here's the honest answer — and how to redact so nobody can read what you hid.
It depends on how it was redacted. Text blacked out with Unicode block characters (like █) cannot be uncovered — the original letters are permanently replaced, so there is nothing hidden underneath to reveal or decode. Image and PDF redaction can sometimes be reversed, but only when the redaction was done badly — a black box drawn over still-present text, or blur instead of deletion.
When you run a sentence through a redacted text generator, each letter is replaced with a block character such as █, ▓, ■ or ⬛. The result is just text — the word agent literally becomes █████. The original letters are not stored anywhere in the output, so:
This is exactly why Unicode redaction is great for joke spoilers, SCP-style entries, and privacy-friendly screenshots: once you copy the blacked-out version, even you can't get the original back from it.
The leaks you hear about — where someone "revealed" redacted text — almost never involve breaking a true blackout. They involve a redaction that never deleted the underlying data:
The takeaway isn't "you can always uncover it" — it's that covering is not the same as removing. If the original data survives anywhere in the file, the redaction is only as strong as someone's willingness to dig.
If your goal is to hide something for good:
And the reverse rule, for anyone hoping to read someone else's redaction: if it was done properly, there is nothing to recover — by design.