Chapter 7 · Reading 6 of 6
Tokenization, masking and steganography
Security+ SY0-801 · Objectives 1.3, 3.3
A campus bookstore cashier needs to refund a purchase. An analyst needs to count sales by hour. Both work with the same transaction, but only one may need a payment capability. The design question begins with what each person must accomplish.
Trace the sensitive value
A processor returns an opaque payment token. The campus bookstore stores the token against the sale. In a vault-based design the provider controls the mapping back to the account number. The token reduces the number of systems that handle the original value. Determine whether possession of the token can still initiate a charge. A token usable for transactions is a sensitive credential even if it reveals no account number.
Masking controls presentation. Showing only the last four digits on a receipt leaves the original elsewhere in the system. Encryption transforms the original under a key. Authorized decryption restores the original without a lookup table. Hashing a predictable value such as a phone number may allow guesses from a small candidate space. Evaluate who has the original data, the mapping, the keys and any auxiliary data that could reidentify a person.
Recognize concealment as a separate property
Steganography conceals the presence of a message within a carrier, such as an image or audio file. Encryption conceals meaning. A hidden message can be readable once discovered, and an encrypted archive can be plainly visible. A watermark may help establish provenance while leaving the main content readable. Choose the property required by the use case before choosing a control.
An organization receives an unexpected media attachment through its document portal. Preserve the original and its digest. Compare its provenance, expected format and purpose with approved submissions. A mismatch between declared and observed file type merits review. File size or unusual image statistics alone cannot prove a concealed message. Apply the authorized content-handling process and record what the evidence establishes. Test privacy claims by following tokens through exports, logs, support tools and backup copies.
A token that still tells a story
Imagine an analyst who sees the same opaque token on every purchase by a bookstore customer. The original card number may be inaccessible, yet repeated values reveal a transaction history. Combine that history with a named support ticket and a person may become identifiable. Before sharing a dataset, inspect stable identifiers, who can join them to other records and what the receiving team actually needs. Aggregate counts may answer the business question.
By the end of the chapter you can describe several distinct assurances: a verified identity, a confidential connection, a controlled decryption decision and a deliberately limited disclosure. The next chapter places those assurances into an architecture where different teams may own each dependency.
Inspect the three records, then use the evidence to answer the question.
1. Checkout
Checkout exchanges the card number for token T6.
2. Warehouse
Warehouse aggregates sales using T6.
3. Refund desk
Refund desk alone can request the original value.
Apply the reading
What would you inspect before calling a token safe to share?
Compare your reasoning
Determine who can map it to the original and whether possession authorizes transactions or links a person's records.