New device transaction rule
A new device transaction rule detects a payment initiated from a device the account has not used before, usually in combination with an amount, a new beneficiary or a recent credential change. A device the account has never used is not suspicious in itself, but it removes one of the signals that normally ties activity to the genuine customer.
The fraud scenario this rule addresses
An attacker with stolen credentials logs in from their own phone or computer, not from the victim's usual device. Because the credentials are correct, the login succeeds, and the only thing that differs from the account's pattern is the device. A payment shortly afterwards, particularly a large one or one to a new payee, is the stage at which the loss occurs.
The same signal appears after a SIM swap, where an attacker has taken over the phone number and registers a new device to receive one-time codes.
Data the rule requires
- Device identifier or device fingerprint for each session
- The account's history of devices and when each was first seen
- Session timestamp, channel and application version
- Transaction amount and beneficiary
- Recent changes to credentials, contact details or enrolled devices
Example logic
Illustrative pseudocode: it compares each event with the rule's parameters and flags the ones that match. Parameters are symbolic: every value is institution-specific and has to be derived from your own transaction distribution.
RULE new_device_transaction
FOR EACH payment P on account A in session S:
known = S.device IN devices_seen(A)
device_age = P.timestamp - first_seen(A, S.device)
IF NOT known OR device_age <= :new_device_period:
IF P.amount >= :min_amount
OR P.beneficiary IS NEW for A
OR credentials_changed(A, within = :recent_change_window):
RAISE alert(rule = "new_device_transaction",
account = A, evidence = S + P)
Choosing a threshold
The parameters are how long a device counts as new, the amount at which a new device matters, and which companion conditions turn a bare new-device event into an alert. A long new-device period is cautious but treats a customer's own recently bought phone as risky for a long time. A short period lets an attacker simply wait.
A rule on new device alone would alert on every phone upgrade, so the companion conditions carry the weight. Requiring a high amount makes the rule precise but means a smaller drain on a new device goes unseen; requiring only a new beneficiary catches the typical takeover sequence but also a customer buying a new phone and paying a friend.
How a device is identified also acts as a threshold. A narrow identifier flags every browser update or cleared cookie as a new device, and a loose one fails to tell an attacker's device from the customer's.
False positives to expect
Customers change devices routinely: a new or replacement phone, a factory reset, a switch between a phone and a laptop, a shared family tablet, a browser update or cleared cookies that change a browser-based identifier, and travel on a borrowed or hired device.
Phone upgrades cluster after product launches, so alert volume spikes predictably. Using the new device together with other changes, rather than alone, and trusting a device once the customer confirms it, removes repeat alerts for the same genuine phone.
Investigating an alert from this rule
- Establish what is new: a device never seen on this account, or one seen on other accounts too. A device linked to several unrelated accounts is a strong warning sign.
- Check how the device was enrolled and whether the enrolment passed the usual verification.
- Check whether credentials, contact details or the registered phone number changed just before the enrolment.
- Compare the session's location and network with the account's usual pattern.
- Review what the session did: browsed, added a payee, changed settings, or paid immediately.
- Contact the customer through a channel on record from before the change, and ask whether they bought or enrolled a new device.
Related rules
Where this rule fits
The use case this rule supports is covered in account takeover, and the fraud rules engine page describes how conditions like these are expressed and versioned.
Common questions
What is a new device transaction rule?
A new device transaction rule is a fraud rule that raises an alert when a payment comes from a device the account has not used before, normally combined with a large amount, a new payee or a recent credential change. A new device alone is common and harmless, but it removes one of the signals that ties activity to the genuine customer.
A rule on the device alone would alert on every phone upgrade, so the companion conditions do most of the work. The way a device is identified matters too: a narrow identifier treats every cleared cookie as a new device, while a loose one cannot tell an attacker's device from the customer's.
The pattern is typical of account takeover, where correct credentials are used from the attacker's own equipment.
Discuss your transaction monitoring requirements
Tell us which transaction flows you need to monitor, which systems Akoben would integrate with, and how your fraud team works today.
Request a demo