- 01Spoofing falsifies the origin, and phishing is the deception that takes advantage of it
- 02In email, authentication protects the exact domain, not the similar one or the visible name
- 03In a call, the number shown on the screen does not prove who is calling
- 04In biometrics, spoofing is called a presentation attack
- 05In remote onboarding, spoofing targets the document, the face, and the device
- 06What VU does
- For NIST (CSRC glossary, from CNSSI 4009-2022), spoofing is an attempt to access a system by posing as an authorized user.
- DMARC, an IETF standard since May 2026 (RFC 9989), only directly combats exact-domain spoofing: it does not cover look-alike domains or the sender’s visible name.
- NIST’s digital identity guidelines (SP 800-63A-4, July 2025) require presentation attack detection in remote biometric capture to mitigate spoofing.
- ENISA (2022) notes that, in remote onboarding, someone impersonating another person can falsify the identity document, the face, or both.
In June 2021, the FBI office in Phoenix warned that a victim’s caller ID could show "FBI-Phoenix" as the caller, and that it was an illusion: on the other end there could be someone anywhere in the world.
In Argentina, the Undersecretariat of Information Technology of the Province of Santa Cruz analyzed in 2020 an extortion email that seemed to come from the recipient’s own account. Its explanation: attackers use "a technique called spoofing, in which they try to disguise the origin of a message".
Both situations are spoofing: falsifying the origin of something so that it appears legitimate. The term covers more than emails and calls. In remote onboarding, a document or a face in front of the camera is also falsified, and each case has a different control with a different limit.
Spoofing falsifies the origin, and phishing is the deception that takes advantage of it
Spoofing is the falsification of the origin of a communication or an access, and, in our reading, phishing is the deception that usually takes advantage of it. Organizations define spoofing from complementary angles. The NIST glossary, which takes the definition from CNSSI 4009-2022, describes it as "An attempt to gain access to a system by posing as an authorized user", and adds that posing as another, masquerading, and mimicking are forms of spoofing.
The FBI defines it by the falsified data: it occurs when contact information (phone number, email, and website) is deliberately falsified to appear to come from a legitimate source.
INCIBE, Spain’s cybersecurity institute, translates it as impersonation and lists its most common forms:
- Website spoofing — a fake website copies the design of the original and may use a similar URL.
- Email spoofing — the address of a trusted person or entity is impersonated in order to ask the victim for information.
- IP address spoofing — the attacker falsifies their address to make it pass as another; according to INCIBE, it is usually used in DDoS attacks.
- DNS spoofing — the DNS of the infected device is modified so that the browser takes the user to another website chosen by the attacker.
The difference from phishing, stated in one sentence, is our reading of INCIBE and the FBI: spoofing falsifies who appears to be on the other side, and phishing is the deception that asks for something. INCIBE says it is used "together with phishing". The FBI defines them separately, but in the IC3 annual report it combines them into a single category, Phishing/Spoofing, in which each complaint has a single crime type.
In email, authentication protects the exact domain, not the similar one or the visible name
Email can be falsified because its original design does not prevent it, and authentication covers only part of the problem. The summary of RFC 7208 says so: "Email on the Internet can be forged in a number of ways." The protocols do not restrict which domain a server can declare as the origin of the message.
Three IETF standards respond to that:
- SPF (RFC 7208) — the owner of a domain authorizes which servers can send on its behalf, and the receiving server verifies that authorization.
- DKIM (RFC 6376) — a cryptographic signature associates the message with the domain that takes responsibility for it.
- DMARC (RFC 9989) — the domain owner publishes what to do with mail that does not pass validation: do nothing, send it to spam, or reject it. It has been an IETF standard since May 2026.
For the domain owner, the policy that matters is rejection. The United States cybersecurity agency (CISA) writes in its BOD 18-01 directive: DMARC in "reject" provides the strongest protection against spoofed email, because the unauthenticated message is rejected at the server before delivery.
With "p=none", however, the spoofed email is still delivered. DMARC reports add a piece of data that otherwise does not arrive, as CISA says: if an attacker spoofs your domain, they are not going to copy you.
The limit is set by RFC 9989 itself, in section 2.2: "DMARC can only be used to combat specific forms of exact-domain spoofing directly." It does not cover visually similar domains or the sender’s visible name, and its section 11.4 leaves out of scope attacks that put a known name in front of any address. Three variants remain, then: the falsified exact domain, which DMARC in reject mode stops; the similar domain; and the false visible name.
Based on its header, the email analyzed by Santa Cruz fits the first variant, according to our reading. That header showed Received-SPF: softfail, and the Undersecretariat explained it this way: "That IP is not authorized to send emails on behalf of the santacruz.gov.ar domain." What the recipient sees in their inbox is in email phishing.
In a call, the number shown on the screen does not prove who is calling
In phone spoofing, the caller ID is falsified, so the number or name shown by the phone does not say who is on the other end. INCIBE describes it as "identity impersonation through the telephone number of a recognized entity", in four stages: identification of the victim, manipulation of the identifier, social engineering, and obtaining the benefit. The sectors most often impersonated, according to INCIBE, are banking, electricity, and telecommunications.
The practical rule is the same on both sides of the Atlantic. INCIBE recommends hanging up and calling the company at the official number listed on its website.
The BCRA, without using the word spoofing, asks in its 2025 recommendations not to use the ATM, the app, or home banking while in communication with an alleged bank representative, to hang up, and to verify through official channels. It also warns that a WhatsApp number can call itself "Help desk" or use a bank’s image.
We did not find a text from an Argentine organization that names caller ID falsification. Other countries regulate it. In Brazil, Anatel maintains measures against "the alteration of the originating number of calls, a practice known as spoofing", and can order the blocking of the interconnections of the responsible operator for up to one month, extendable to three in the event of recurrence.
In the United States, federal law prohibits transmitting a misleading identifier when there is intent to defraud, cause harm, or obtain something of value.
In biometrics, spoofing is called a presentation attack
In identity verification, spoofing of a face or a fingerprint is what the standard calls a presentation attack: showing the sensor an artifact so that it takes it for the person.
The summary of the 2016 edition of ISO/IEC 30107-1 says it with the market’s word: biometric data is used to create "spoofs (or fakes)", and the presentation of a "biometric spoof", such as an image or video on a tablet or a silicone or gelatin fingerprint, is detected with presentation attack detection methods, or PAD.
NIST makes it a requirement. Its SP 800-63A-4 guidelines require that, when biometrics are collected and compared remotely, the provider implement PAD "to confirm the genuine presence of a live human being and to mitigate spoofing and impersonation attempts", with tests compliant with ISO/IEC 30107-3:2023.
The injection attack is another one. NIST describes it as inserting modified or false media between the capture point and the element that performs the comparison, and ENISA clarifies that an injected video is not a presentation attack instrument, because it is not shown to the camera: it bypasses it.
NIST adds a nuance. Live capture and PAD provide some protection against injection, because they make it more difficult, but they are not enough for all cases.
In remote onboarding, spoofing targets the document, the face, and the device
In remote identity verification, spoofing can target three things, and each one requires its own control. ENISA writes in its report on remote identity proofing: someone seeking to deceive the system by impersonating another person "has the choice of spoofing" the identity document, the face, or both.
- The document — the control seeks reasonable certainty that it is a real, authentic, and valid document issued by a government, and not a forgery. In documents with a chip, the public key infrastructure authenticates the stored data and photo.
- The face — it is attacked in front of the camera, with a mask or a video on another screen, or by bypassing it with an injection.
- The device — NIST requires the provider to implement technical controls to increase confidence that the media comes from a genuine sensor, for example detecting a virtual camera, an emulator, or a jailbroken device.
The location can also be falsified. NIST defines spoofing of the satellite positioning signal, and Android tells an application whether a location is simulated. Android itself warns that the user may have a good reason to simulate it, so a simulated location is a signal that must be weighed with others.
When the attacker uses a real person’s identity to open a new account, the case is identity theft.
What VU does
Verify, VU’s identity verification, applies facial liveness detection against presentation attacks.
ISO/IEC 30107-3 defines the method, the metrics, and how a presentation attack detection test is reported. The attack sophistication levels are not defined by the standard: they are defined by the iBeta testing program, which applies that methodology and orders them by what it takes to build the instrument, in time, expertise, and equipment. There are three.
The scope of the standard is presentation attacks, those that occur in front of the capture device; injection attacks are outside its scope and are covered with different controls. What VU has evaluated is facial liveness detection, at Level 3 of that program.
The VU capability that provides context about the device is Protect against fraud, with recognition of the equipment used to operate based on browser and mobile signals. The decision on what to do with that signal, and the monitoring of the operation, are resolved in another layer that VU does not provide.
Spoofing always falsifies an origin, and each origin, from the sender to the face, is checked with a different control.
schedule a conversation.
