Cryptographic erase (crypto-shredding)
Sanitising encrypted data by destroying every copy of the key that protects it, so the ciphertext left on the media or in a cloud service can no longer be decrypted.
Cryptographic erase (CE) sanitises data by destroying the key rather than the data itself. If the data was encrypted when it was written, removing every copy of the key leaves only ciphertext that cannot practically be decrypted. NIST SP 800-88 Rev. 2 (2025) classes it as a purge technique, performed by sanitising the keys that protect the data, and it is commonly built into self-encrypting drives. In cloud services, where the customer cannot reach the physical disks, Rev. 2 notes that it may be the only viable purge option, and the same idea is often called crypto-shredding there.
CE is fast, leaves the device reusable, and suits flash media where overwriting is unreliable because of wear levelling. Its strength rests on conditions that have to be checked. The data must have been encrypted from the start, since anything written in plaintext before encryption was enabled can remain. Every copy of the key must be gone, including backups and any key escrow copy, and Rev. 2 recommends zeroization as the way to sanitise the keys. The encryption and the vendor’s implementation must also be trustworthy. For data that must stay confidential for decades, Rev. 2 cautions that CE may not be acceptable, because the ciphertext remains on the media and future cryptanalysis could expose it.
Exam relevance: a scenario is likely to involve disposing of SSDs, or leaving a cloud provider where physical access to the media is impossible. Candidates are expected to recognise cryptographic erase as a fitting answer there, rather than degaussing, which does not affect flash, or clearing, which can miss areas outside user-addressable storage.