What does an NFC door reader that accepts only a card's static identifier prove, and what does a card holding a secret key add?
answer
- a number, not a secret
- possession versus presentation
- fresh challenge from the reader
- key never crosses the air
- mutual: both sides prove
basics
~20 sA static identifier is a value anyone who reads it can present again, so it proves nothing about possession. A card with a secret key answers a fresh reader challenge, which proves it holds the key.
solid answer
~40 sA reader that accepts only a static identifier learns that some device presented that number; it learns nothing about who holds the card or whether the number was copied, because the identifier is not a secret and can be reproduced on another tag or an emulating device. NIST SP 800-98 treats cloning as an intentional attack on an RFID business process and says authenticity checks typically need cryptography and anti-cloning mechanisms. A card with a secret key instead answers a random challenge from the reader, so a recorded answer does not fit the next challenge, and with mutual authentication the card challenges the reader too. The residual risks are key secrecy and key management at the readers, and relay of a genuine exchange, which this does not address.
go deeper
Recall that a static identifier is a number that can be presented again, so it identifies but does not authenticate; a secret-key answer to a fresh challenge does.
Explain challenge-response, why the answer is useless later, and what mutual adds; name key management and key extraction as the cost NIST lists.
Review an access system: which readers accept bare identifiers, whether a dual-accept fallback stays open, where keys are stored, and what relay still gets through.
Frame the credential choice as key custody and revocation across readers and owners, not as a card brand, and fund compensating controls for doors that cannot migrate.
## What an identifier proves Many NFC-A and NFC-B cards (the ISO/IEC 14443 family) expose a fixed identifier when a reader selects them. If the access system looks that value up in a table and opens the door, the entire credential is a **public number**. The reader can only conclude that a device presented that value, not that the device is the card the value was printed on. NIST SP 800-98 defines a cloned tag as a tag made to be a duplicate of a legitimate one, created by reading data such as an identifier from a legitimate tag and writing it to a different tag. It lists cloning as an intentional attack on an RFID business process, with documented examples in financial payment and access-control tags. Its tag-threat list also names replacing a tag with another one. The guide's point for a designer is that identification and authentication are different jobs: a static identifier does the first and not the second. ## What a keyed credential adds A credential that holds a secret key changes the question the reader asks. Instead of "what number are you?", the reader asks "prove you hold the key". NIST describes the pattern in its discussion of keyed-hash authentication: the tag computes and returns a keyed hash over a random challenge supplied by the reader, and mutual authentication is possible when tag and reader each provide a challenge to the other. Design-level properties that follow: - **Possession is demonstrated, not asserted.** The key never crosses the air; only a value derived from it and the challenge does. - **A recorded answer is useless later.** The next challenge differs, so an earlier answer does not fit. (Replay and freshness as a general subject are covered with network attacks; here it is one clause.) - **The reader is checked too.** With mutual authentication a counterfeit reader cannot make the card talk as if it were genuine, which protects the card side as well. - **Integrity comes along.** NIST notes that cryptographic authentication transactions often provide integrity for the data in the exchange. Card families built on this design, such as MIFARE DESFire with its AES mutual authentication, work on the same principle: the card proves possession of a secret key. The sources used for this answer do not document that card, so check the chip's own data sheet for exactly what it protects. ## Comparison | Property | Static-identifier card | Card with a secret key | |---|---|---| | What the reader receives | A fixed value | An answer to a fresh challenge | | Copy of what was read is accepted later | Yes | No, for a recorded answer | | Counterfeit reader detected by the card | No | Yes, with mutual authentication | | Needs key management at readers | No | Yes | | Stops relay of a live exchange | No | No | ## What it costs and what it leaves open NIST is explicit about the weaknesses of keyed schemes, and a reviewer should repeat them: 1. **Key management.** Managing the keys poses challenges similar to password management and may be impractical if readers lack reliable access to a key management system. 2. **Authenticity holds only while the key is secret.** If an adversary has physical access to a tag and can obtain its key, the tag can be cloned; NIST adds that this takes expertise in reverse engineering and in producing a facsimile. 3. **Shared keys rely on trust.** When keys are shared across organisations, authenticity claims rest on an implicit trust between them. 4. **Cost.** NIST notes that keyed-hash authentication needs more computing power, and so more complex tag designs, than a password comparison. 5. **A signed value stored on the tag is not enough.** NIST warns that digital signatures not generated by the tag itself are subject to replay, because a previously obtained signature can be reproduced on a cloned tag; the tag has to answer something fresh. 6. **Relay is untouched.** A genuine card and a genuine reader can still be made to act as if adjacent; mutual authentication does not bound distance. ## What to look for in a review - **Inventory:** which door controllers accept a bare identifier, and whether any reader still accepts both a keyed and a static credential, which keeps the weak path open. - **Evidence:** the reader configuration showing what it accepts, and where the keys live. - **Honest wording:** "the badge has an ID" is a statement about identification; "the badge cannot be duplicated" is a claim the design must earn with a key. - **Compensating control:** for readers that cannot move yet, add a second factor at the doors that matter. ## Interview framing A strong answer names the distinction first (identifier versus proof of possession), states the mechanism (fresh challenge, secret key, mutual), and ends with the honest limits (key custody and relay). A weak answer says the card is "encrypted" without saying what the reader verifies.
- A vendor says its static-identifier card is 'encrypted'. What do you ask?Ask what the reader verifies. If the reader only looks up an identifier, encryption of the radio link does not make the card unclonable, because the value is still presented, not proven. Ask whether the card answers a fresh challenge with a secret key, who holds that key, and whether the reader or the card can be counterfeited without it.
- Does mutual authentication also stop a recorded exchange being used later?A recorded answer to an earlier challenge does not fit a new random challenge, so that replay fails. The card also refuses a reader that cannot prove the key. It does not stop a live relay, because in a relay the genuine card answers the genuine challenge in real time.
A name written on a visitor sticker can be copied by anyone with a pen; asking the person to sign fresh in front of the guard shows they hold the pen that the signature card was made with.
- Says a card ID number is a secret credential
- Treats radio-link encryption as proof the card cannot be cloned
- Claims mutual authentication also prevents relay
- Ignores where the readers keep the keys
- Calls a signed value stored on the tag replay-proof