46 terms that onboarding, login and fraud teams use, from account takeover and SIM swap to money mules and coached payments. Each links to related terms, to its module where DeepID has one, and to its source where a regulator defines it.
Aadhaar face authentication compares a live face with the image captured at Aadhaar enrolment, one to one, with the resident’s consent. UIDAI recorded 28.29 crore face authentications in November 2025.
In DeepID’s one-tap login it is the consent route in the Aadhaar app, after the network check and the SIM check and before a one-time code.
Aadhaar offline KYC is a digitally signed copy of a resident’s Aadhaar details that the resident downloads and shares with consent, so a bank can verify UIDAI’s signature and detect tampering without querying UIDAI online.
It is one of the forms of eKYC DeepID verifies at sign-up, with DigiLocker, PAN and CKYC.
Account takeover is when someone other than the customer gains control of their account, usually with phished credentials, a forwarded one-time password or a swapped SIM.
The person and often the device are wrong, so authentication and device signals can catch it.
AML screening checks a customer against sanctions, politically exposed person and adverse-media lists at onboarding and after it, with a possible match reviewed by a person rather than refused on a name alone.
It sits beside transaction monitoring and reporting, which stay with the bank’s own systems.
Source: RBI, Feb 2016, updated Aug 2025 (opens in a new tab)
RelatedeKYCKYBMoney mule
The anti-fraud stack is onboarding, login and fraud intelligence on one SDK, so each step of a customer’s journey is checked against everything before it.
It is the category line DeepID uses for itself: verification at sign-up, one-tap login, and the session read before every payment.
An authorised push payment scam is one where the genuine customer is deceived into sending the money themselves.
The customer is real, so each authentication check passes on its own. The session around the payment shows the scam: the call, the shared screen and the payee.
RelatedCoached paymentDigital arrest scamSession fraud intelligence
Behavioural biometrics compare how a person types, taps and moves through an app with how the account’s owner usually does.
On its own it can be imitated by malware; combined with device integrity and call context it is much harder to fool.
CKYC is India’s Central KYC Records Registry, where a customer’s KYC record is held once and reused by regulated entities instead of being collected again.
RBI’s KYC Master Direction defines the registry and requires records to be filed with it. DeepID checks a CKYC number against the registry at sign-up.
Source: RBI, Feb 2016, updated Aug 2025 (opens in a new tab)
RelatedeKYCPANDigiLocker
A coached payment is made by the real customer while a scammer instructs them, usually on a phone or video call, sometimes with the screen shared.
The signs are in the session: an active call, dictated and hesitant typing, an unusual amount and a first-time payee.
RelatedAuthorised push payment scamDigital arrest scamRemote access tool
A cooling-off period limits what can be done for a set time after a risky change, such as a new device, a new SIM or a new payee.
Singapore’s Shared Responsibility Framework includes a 12-hour cooling-off after a new device or token is activated.
A deepfake is synthetic video, audio or imagery made with AI to look or sound like a real person.
In banking it appears as generated selfies at onboarding, cloned voices on calls and faked colleagues on video. FinCEN alerted US financial institutions to the pattern in November 2024.
Device binding links a customer’s account to a specific, registered device, so that a new device has to be verified before it can act.
It is usually paired with SIM binding and a cooling-off period after a change.
RelatedSIM bindingCooling-off periodDevice integrityDevice fingerprint
A device fingerprint is a device ID that survives reinstalls and factory resets, so a bank recognises the same phone across sessions and sees one phone behind many accounts.
It is the identity that device binding ties the account to, and that the mule graph links accounts through.
Device integrity checks whether the phone and the app can be trusted: not rooted or jailbroken, not an emulator, not hooked or cloned, and not under remote control.
An attack on the device usually leaves a trace here even when the credentials are correct.
DigiLocker is the Government of India’s document wallet, from which a resident shares documents issued to it, such as Aadhaar and PAN, with consent.
RBI’s KYC Master Direction accepts an equivalent e-document issued to a customer’s digital locker account in place of the paper original. DeepID reads such documents at sign-up.
Source: RBI, Feb 2016, updated Aug 2025 (opens in a new tab)
RelatedeKYCAadhaar offline KYCPAN
A digital arrest scam is an impersonation scam common in India, where callers posing as police, customs or regulators keep the customer on a video call and order them to move money to a “safe account”.
The customer pays from their own phone with their own credentials, so each check passes on its own. The call that runs across the payment, and the payee, show the scam.
RelatedCoached paymentAuthorised push payment scamMoney mule
eKYC verifies a customer’s identity against the record that issued it, such as Aadhaar through offline KYC or DigiLocker, PAN or CKYC, instead of a photocopy.
RBI’s KYC Master Direction sets out the forms it may take in India, including digital KYC and video-based customer identification.
Source: RBI, Feb 2016, updated Aug 2025 (opens in a new tab)
Face match compares the live face from a liveness check with the portrait on Aadhaar or on the document, one to one.
Face search is the one-to-many form: the same face looked for across a bank’s customers, so one face behind many accounts is caught.
RelatedLiveness detectionAadhaar face authenticationSynthetic identity
The Financial Fraud Risk Indicator is a risk grade, medium, high or very high, that the Department of Telecommunications’ Digital Intelligence Platform gives a mobile number from reports to I4C, DoT’s Chakshu and banks, launched in May 2025.
Banks and UPI apps use it to add checks when a payment goes to a flagged number.
A fraud risk management system scores and monitors transactions, holds alerts for review and manages fraud cases.
It sees the payment once it exists; it rarely sees the session in which the payment was created.
An injection attack feeds a fake image or video stream directly into an app’s camera pipeline, through a virtual camera, an emulator or a hooked app.
Because it bypasses the camera, it can pass a liveness check built for printed photos and masks. CEN/TS 18099 sets out how injection attack detection is tested.
The journey builder composes onboarding, login, recovery and payment journeys from DeepID’s modules, with the checks, the order and the fallbacks the policy sets, versioned per market.
A journey changes without an app release, and every run is logged with its decisions and reasons.
Know your agent is the practice of confirming that an AI agent acting for a customer is declared, tied to a verified customer and device, and acting within a mandate the customer set.
It separates a customer’s approved assistant from automation that only pretends to be one.
KYB, know your business, verifies a company or merchant against GST and company registries, and its directors and beneficial owners like any customer.
It is how shell merchants, and the people behind them, are kept off a payment network. RBI’s KYC Master Direction sets out due diligence for legal entities, including beneficial owners.
Source: RBI, Feb 2016, updated Aug 2025 (opens in a new tab)
RelatedeKYCAML screeningMoney mule
Liveness detection checks that a face in front of the camera belongs to a live person, not a photo, mask or screen. ISO/IEC 30107-3 is the standard for testing and reporting it.
It proves someone is there. It cannot tell whether that person is being coached, and an injected stream can reach it unless the device is read with it.
The Mobile Number Revocation List is DoT’s list, on its Digital Intelligence Platform, of mobile numbers disconnected for fraud or misuse.
In January 2025 RBI directed banks and other regulated entities to use it to clean their customer records and to watch accounts linked to revoked numbers for mule activity.
A money mule is an account, and often a person, used to receive stolen money and pass it on.
Mule accounts are recruited, rented or opened with synthetic identities, and each one looks normal on its own.
A mule network is a ring of accounts linked by shared devices, SIMs, faces and payees, used to layer and move scam proceeds.
The ring shows only when accounts are linked; a single-account view sees nothing unusual.
MuleHunter.AI is a mule-account detection model built by the Reserve Bank Innovation Hub, which the Finance Ministry asked banks to adopt in December 2024.
DeepID’s payee and mule graph works alongside shared utilities such as this and claims no integration with them.
Source: Outlook Business, Dec 2024 (opens in a new tab), press report
NFC chip verification reads the chip in an ePassport or eID and authenticates it cryptographically, so the data page cannot be forged on its own: the chip has to agree with it.
ICAO Doc 9303 specifies the machine-readable zone and the chip that DeepID reads.
Source: ICAO, 8th edition, 2021 (opens in a new tab)
RelatedeKYCFace match
One-tap login signs a customer in by confirming their number with the strongest check the phone and network allow, so most customers type nothing.
In DeepID’s order: silent network verification, then the SIM check, where the phone sends one text from its own SIM, then Aadhaar consent, then a one-time code as the last fallback. The first that verifies ends it, and the bank’s server redeems a single-use token before signing anyone in.
RelatedSilent network verificationSIM bindingRisk-based authenticationAadhaar face authentication
PAN is the Income Tax Department’s identifier for a taxpayer, and the number a bank checks against the issuing record at sign-up under RBI’s KYC rules.
A PAN already on another customer’s record is a sign of a reused or synthetic identity.
Source: RBI, Feb 2016, updated Aug 2025 (opens in a new tab)
RelatedeKYCCKYCSynthetic identity
Payee risk is an assessment of where money is going: how old the payee is, what accounts and devices it is linked to, and whether it sits in a mule cluster.
It matters most at the moment a payee is added and the first payment is made.
Reason codes are short, readable labels that explain why a decision was made, such as ACTIVE_CALL or SIM_MOVED.
The same codes serve the analyst, the customer conversation and the audit trail.
RelatedFraud risk management systemSession fraud intelligenceJourney builder
A remote access tool lets someone else see or control a phone, through screen sharing or an installed app.
Scammers use it to watch the customer enter details, or to drive the banking app themselves.
A retrospective is DeepID’s free first step for fraud teams: a read of 30 to 90 days of a bank’s own transactions and closed cases that shows the payee and mule links in them.
It cannot show session signals, such as an active call or a shared screen, which exist only once the SDK runs; those are proved in a shadow proof.
Risk-based authentication adjusts the checks a customer faces to the risk of the moment, adding a step-up only when the session calls for it.
India’s Authentication Directions, in force since 1 April 2026, keep two factors as the baseline, keep SMS OTP as a valid factor and allow risk-based checks beyond them.
Session fraud intelligence is DeepID’s Intelligence family: what a bank reads around a login, a new payee or a payment in the minutes before the payment exists. The device and its integrity, how the customer types, whether a call or a shared screen is running, and where the money is going.
Your FRM sees the payment once it exists. Session fraud intelligence is read before, which is where scams, takeovers and deepfakes leave their traces.
RelatedFraud risk management systemReason codesAuthorised push payment scamAnti-fraud stack
A shadow proof is the step after the free retrospective on DeepID’s proof path: the SDK runs on live traffic for six to eight weeks and scores it without acting, so a bank compares its decisions with its own outcomes before any customer is affected.
It is where session signals are proved, and where a fraud team sees how many genuine payments each policy would have held.
Silent network verification asks the mobile network, in the background and over mobile data, to confirm that a phone number belongs to the SIM in the phone making the request, with nothing to type.
There is no code, so there is nothing to phish or forward. It confirms the number; who is holding the phone is the question SIM and device binding answer, which is why the two are paired.
RelatedOne-tap loginSIM bindingSIM swapRisk-based authentication
SIM binding ties an account or app to the SIM registered for it, and checks again whenever the SIM is removed, swapped or moved to another phone.
UPI apps already bind to the SIM. In November 2025 India’s Department of Telecommunications directed messaging apps to work only while the registered SIM is present and active, with a deadline reported as 31 December 2026.
Source: The Hacker News, Dec 2025 (opens in a new tab), press report
A SIM swap moves a customer’s phone number to a new SIM, often by tricking or bribing the mobile operator.
Once the number moves, one-time passwords go to the fraudster.
RelatedSIM bindingSilent network verificationAccount takeoverMobile Number Revocation List
Step-up authentication asks for an extra check, such as a face check on the bound phone, only when a session or action carries more risk.
Good customers see it rarely; risky sessions see it at the right moment.
A synthetic identity combines real and invented details, and increasingly AI-generated documents and faces, to create a person who passes KYC.
It shows up as the same device, face or SIM behind many applications.
UPI is India’s real-time payment rail, run by NPCI. Onboarding binds the UPI app to the device and needs an active SIM linked to the bank account.
That is why SIM binding is already familiar to Indian customers, and why a payment is checked against the SIM and phone bound before it.
Video KYC verifies a customer on a live video call, agent-led or self-serve, with the face, the document and the location captured during it. RBI’s KYC Master Direction describes the agent-led form as V-CIP, video-based customer identification.
DeepID runs the same injection and deepfake checks on the video stream as on a selfie, so a pre-recorded or generated face does not reach the agent as a live one.
Source: RBI, Feb 2016, updated Aug 2025 (opens in a new tab)
We show your fraud, security and engineering leads how DeepID would run your sign-up, login or payment journey, then plan a pilot on your own traffic.