Blog

Tap, chip, swipe or phone wallet: how each protects your card data

Four ways to pay compared on what the merchant receives, what a thief could copy and where the remaining risk sits, from magnetic stripe to phone wallet.

Every way of paying with a card sends some information to the merchant, and the real question is what that information is worth to a thief who copies it. A magnetic-stripe swipe sends the same data every time. A chip creates a new code for each purchase. A tap sends that same chip code over a short radio link. A phone wallet adds a stand-in number so the merchant never sees the real one. Those four design choices explain most of the security difference between the ways to pay, and they also show where each method still has a weak point. This is a comparison, not a ranking. Each method suits some situations and fails in others.

The question to ask of each method

For each way to pay, ask four things:

  1. What does the merchant receive? The real card number, a stand-in, or something in between.
  2. What can be copied, and is the copy useful? Data that is the same every time can be reused. Data that changes cannot be.
  3. What stops a stolen or lost card or device being used? A PIN, a phone unlock or nothing.
  4. Where does the remaining risk sit? Every method moves risk without removing it.

The answers, in a table, are below. The rest of the article explains each row.

Swipe (magnetic stripe) Chip (inserted) Tap (contactless card) Phone wallet (tap or online)
What the merchant receives The real card number and static stripe data Real card number and a one-time code for this purchase Real card number and a one-time code for this purchase A device-specific token and a one-time code, not the real number
What a copy is worth A copy can make a counterfeit card A copied transaction cannot make a working card-present counterfeit Same as chip A token has little value away from that device and merchant
Check on the person Signature, sometimes none PIN or signature, depending on card and terminal Usually none for small purchases Phone unlock (face, fingerprint or passcode)
Main remaining risk Skimmers and counterfeit cards The number used online after a breach; a lost card A lost card used for small purchases; the malware relay scam A weak phone lock; being tricked into adding a card to a criminal's wallet

Swipe: the oldest design and the easiest to copy

A magnetic stripe stores data in a pattern that a reader decodes. The information on the stripe is static: it is the same each time the card is swiped. Anyone who reads it, with a legitimate terminal or a hidden skimmer, holds a complete copy that can be written to a blank card. A counterfeit card made that way can be swiped at a terminal that accepts stripes, and the terminal cannot tell the copy from the original.

That is the weakness chip cards were designed to fix. The magnetic stripe still exists on most cards in the United States, as a fallback when a terminal cannot read the chip. Some terminals, especially older unattended ones, still accept it. Two dates in the sources are worth noting. According to trade press, the shift of liability for counterfeit card-present fraud to the party that had not adopted chip came in October 2015 for most U.S. merchants. Automated fuel dispensers were delayed, first from October 2017 to October 2020 and then to April 17, 2021. Unattended fuel pumps are a commonly cited place for skimmers, though the sources used for this article give no incident counts. If you want to reduce that risk, look at the reader and the card slot for anything that seems added, use a pump in view of the station, and prefer chip or a wallet where the terminal offers them.

When does a swipe still make sense? When it is the only thing the terminal accepts, or when a chip fails and the terminal offers it as a fallback. It is also the method with the weakest protection, so it deserves a closer look at the statement afterward.

Chip: a new code for every purchase

An EMV chip is a small computer in the card. When you insert it, the terminal and chip exchange messages. The chip produces a cryptographic code, called a cryptogram, built from data unique to the card and to that transaction. The issuer checks the code. A copy of an earlier transaction's data cannot be replayed as a new purchase, because the new purchase needs a new code.

This is why the counterfeit card problem shrank where chip was adopted. Visa says EMV cuts counterfeit card fraud by more than 80 percent compared with the magnetic stripe, a figure from the network's own material. It was designed for exactly the attack that a skimmer performs.

What chip does not fix is the real card number. The number is still sent to the merchant's system in many setups, and if the merchant's records are breached, the number can be used where the chip is not needed: online, by phone and in other card-not-present settings. That is card-not-present fraud, and it is why fraud moved there as chip arrived. Chip also does not stop a lost card from being used, though a PIN, when the terminal asks for one, does.

Tap: the same chip, carried by radio

A contactless card holds the same kind of chip and sends the same kind of one-time code, over a short-range radio link of a few centimeters. The security design is therefore about the same as inserting the chip. The convenience changes the habits: fewer terminals are touched, the card never leaves your hand and a purchase is quick. The contactless payment glossary entry gives the definition.

The remaining risks are concrete. First, a lost or stolen contactless card can be used for small purchases without a PIN at many terminals, so a lost card should be reported at once. The $50 federal cap on liability for unauthorized use applies, and reporting a lost or stolen card before it is used means no liability. Second, there is the malware relay scam called ghost tapping, in which a victim is tricked into installing an app and a criminal's device replays the card data. Can a stranger steal from your card by tapping it in a crowd? Myths tested tests the popular claims about it, including the crowd-scanning story that the sources do not support. For what happens inside a tap, from the terminal's request to the approval, see what happens in the seconds after you tap: network, issuer and merchant.

Phone wallet: a stand-in for the number

A mobile wallet stores a version of your card on the phone. When you add the card, the issuer creates a token linked to the card and to that device. When you pay, the wallet sends the token and a one-time code, and the merchant never receives your real number. This is tokenization. A merchant that is breached holds a token that is of little use elsewhere, because it works only with that device and under particular conditions.

A wallet also adds a check on the person. The phone must be unlocked, by face, fingerprint or passcode, before it will pay. A lost phone with a good lock is much harder to use than a lost card. The risks are different:

Setting up the wallet well matters, including which card is the default and what locks are on. phone wallet setup checklist: default card, locks and what to switch on covers that.

A note on cards stored with merchants

Much of today's card use is not at a terminal at all. A streaming service, a rideshare app or an online store keeps your card on file and charges it without your presence. That stored number is the same real number that was typed in once, and it sits in the merchant's records. It is the reason a merchant breach can matter years after a purchase, and also the reason a replaced card number means a round of updates, as Got a replacement card after fraud? A checklist of accounts to update explains.

Some of the protection described for wallets extends here. When a merchant stores a card that you added through a wallet, the merchant may hold a token and not the real number, and card networks have for years offered tokens to merchants for stored cards. Whether a particular merchant does so is not something a customer can see, so treat stored cards as the place where the older risk survives. The practical habits are small: keep stored cards to merchants you use regularly, remove cards you no longer need, and watch statements for charges from places you have not used lately.

What happens when something goes wrong, by method

A short table of events and outcomes helps show where each design earns its keep. The table is qualitative, because the sources used here do not give per-method fraud rates.

Event Swipe Chip Tap Phone wallet
A skimmer at a terminal copies the stripe The copy can become a counterfeit card Chip data cannot make a working counterfeit; a stripe fallback can still be read Not applicable to a skimmer in a card slot Not applicable, since the real number is not sent
The merchant's records are breached Real number exposed Real number may be exposed Real number may be exposed Token exposed, of little use elsewhere
The card is lost Usable where a signature is not checked Usable where no PIN is requested Usable for small purchases at many terminals Needs the phone unlock
A fake "bank" call asks you to install an app Not relevant Not relevant Relay scam possible Relay scam possible
You type the number into a fake website Number exposed Number exposed Number exposed Number exposed, unless a virtual card number was used

The last row is the one that no tap, chip or wallet changes. A person who types a real card number into a fake page has given the number away. how a virtual card number works, and when it beats your real number explains one tool for that case, a separate number for a merchant or a single use.

Reading the comparison for your own situation

The table does not tell you what to do, because the right method depends on the setting. A few labeled illustrations show how the reasoning goes.

  • A busy gas pump at night. A skimmer is a realistic worry at an unattended pump. A phone wallet or a chip insert sends no reusable stripe data, so either suits the situation better than a swipe. A pump inside the view of an attendant reduces the odds further.
  • A small shop abroad that asks you to swipe. The terminal may only take the stripe. If you must swipe, check the statement more closely over the next days, and consider whether a card you can freeze from the app should be the one you carry.
  • A large online order. The method at the till is irrelevant. The risk is where the number is typed, so the question is the merchant and whether a virtual number is available.
  • A crowded commute. Tapping a card or a phone is quick. The documented risks are a lost card and a scam call, not a stranger nearby. Lock your phone, set alerts and report losses fast.
  • A phone that is old and cannot update. Wallet security depends on the device, so a weaker phone may be less suited than a card with a chip.

None of these is a recommendation to apply for anything. They show how to read the trade-offs, and your issuer's own security pages describe the features of your card.

What none of the methods does

Three things are worth remembering for any method.

  1. None prevents you from being persuaded. A scam payment is made by you, and the chip, the token and the lock were all working as designed. The distinction between a charge you did not make and a payment you were tricked into making is covered in scam payment or unauthorized charge: why your protections can differ.
  2. None replaces a look at the statement. Alerts and a habit of scanning the lines catch what the technology does not.
  3. None changes the legal protection. The liability cap of $50 for unauthorized use, and the 60-day written-notice rule for billing errors, apply however the card was used. The CFPB takes complaints if an issuer does not handle a claim fairly.

If you find a charge you did not make, call the number on the back of the card. Do not use a number from a message.

Key takeaways

  • The security difference comes from what is sent. A swipe sends static data that can be copied, a chip or tap sends a one-time code, and a wallet sends a token in place of the real number.
  • Chip and tap use the same chip standard. Their remaining risks are a lost card, the real number being used online after a breach, and the relay scam that starts with a call or message.
  • A phone wallet adds a phone unlock and keeps the real number from the merchant, but it depends on a good lock and on refusing to read verification codes to anyone who contacted you.
  • No method protects the number you type into a fake website. A virtual card number or a close look at the merchant addresses that gap.
  • Whatever method was used, an unauthorized charge on a credit card carries a $50 federal liability cap and a 60-day written-notice rule for disputes.

Terms used in this article

Latest on this topic