Your face is a terrible secret. It turns up in holiday photographs, work profiles and the small circle beside your name on a video call. A business that treats that public image as proof of identity has an awkward problem: somebody else can bring a copy. iProov’s answer is to ask the face to do something a photograph cannot readily do. Be here, under this particular light, at this particular moment.
That idea puts the company in an unusually intimate corner of cybersecurity. A customer looks into a camera; software makes a judgment about whether to trust the encounter. The result can help somebody establish an account, recover access or complete an identity check without visiting a counter. For the organization, the attraction is keeping the convenience of remote service while making impersonation harder.
A phone screen becomes the challenge
The distinctive part is called Flashmark. During a Dynamic Liveness check, the device screen illuminates the user’s face with an unpredictable color sequence. Captured imagery travels to iProov’s cloud, where analysis examines the response. The light ties the capture to a fresh session. A recording from yesterday should not satisfy today’s challenge merely because the face looks correct.
There are three separate questions inside that encounter: does the face match the claimed identity, is it a real person, and is that person present now? It is tempting to collapse them into “facial recognition.” Doing so misses the point. A system can recognize the correct face in a photograph while accepting entirely the wrong evidence about who is holding the device.
Conceptual illustration of iProov’s identity-assurance approach.
The company also offers Express Liveness for lower-risk uses. Its developer materials distinguish that option’s real-person assurance from Dynamic Liveness’s additional emphasis on presence now. Organizations can therefore choose a check to suit the transaction. Recovering an account deserves a different discussion from an ordinary return visit.
One of the more human details lives in the interface. iProov developed an abstract rendering of the user’s face after observing that a mirror image could tempt people to fix their hair and fuss with their appearance. The drawing gives them something to align without quite inviting a grooming session. Even serious security software has to negotiate with vanity.
The fraud came before the face scan
Andrew Bud’s explanation of the company’s origins begins in mobile payments. In a tenth-anniversary interview, he recalled running a business caught up in a large fraud affecting millions of people. The experience redirected his attention toward the identity problem underneath the transaction.
“I realized that the key to this was about authentication and identity.”
Andrew Bud · Founder and CEO · 2021
He founded iProov in 2011. The early difficulty was persuading others to back an unproven technique: Bud later singled out the first Innovate UK grant in 2013 as a breakthrough. Winning the National Cyber Security Centre’s Cyber Den competition in 2017 helped give the company visibility with government buyers. Those are useful milestones for a security supplier. A clever demonstration has to become something institutions are willing to depend on.
The mission remains broad: establish enough trust for people to use services safely online. The actual work is narrower and more demanding. It requires expertise in biometric matching, machine learning, capture design and attack detection. iProov’s published hiring includes red-team and security-engineering roles alongside machine-learning work. The people building the check must understand how someone would try to fool it.
The buyer is the institution. The user is you.
iProov’s customer list includes UBS, ING, Rabobank, the UK NHS and Home Office, GovTech Singapore, the Australian Taxation Office and the US Department of Homeland Security. These are settings where an identity mistake can affect access to money or public services. The company sells to the institution; the individual encounters its technology inside a larger journey.
Singapore illustrates why these checks belong in a broader system. GovTech’s current description of Singpass lists face verification alongside app-based login and SMS authentication, supporting access to government and business services. In June 2026, it added a passkey beta. Different methods coexist because proving identity, logging in and recovering access involve different circumstances. A face check does not have to carry the whole system on its own.
Commercially, iProov supplies cloud software through organizational integrations and partners. Its SDKs cover web and mobile environments, including iOS, Android, Flutter and React Native. Developers put the capture experience into a service, then connect it to backend identity decisions. The customer is buying a working verification capability and ongoing defenses, with implementation work around it.
The clearest public capital figure is the $70 million growth investment from Sumeru Equity Partners announced in January 2022. iProov said the money would support international expansion, particularly in the United States, along with hiring, partners and technology development. That is expansion capital, not a price tag for installing a face scan.
Daily transactions surpassed in 2025, according to iProov’s March 2026 announcement. Transactions count checks, not distinct people.
A biometric corridor meets a changing station
For a more tangible view of the product, consider a railway platform. In July 2023, Eurostar rolled out SmartCheck at London St Pancras for eligible premium travelers. Passengers enrolled their passport, face and ticket before arriving, then used a dedicated biometric corridor. iProov supplied the facial biometric technology; Entrust supplied identity orchestration and digital travel credential technology.
The attraction was moving work out of the station. SmartCheck replaced ticket-gate check-in and the manual UK exit check performed by Eurostar staff. Baggage inspection and French passport control remained. Its trial had received Department for Transport support through an Innovate UK competition. This was a specific redesign of the departure process, with particular passengers and particular checks in scope.
Then the setting changed. SmartCheck ceased operating in February 2025 as Eurostar prepared station changes for the EU’s Entry/Exit System. iProov’s closure notice said the lane would become additional priority space and pointed to new passport gates and booths. The announcement described the technology as successful while explaining why the service was ending.
It would be unfair to call that an algorithm failure. It would be equally misleading to keep presenting SmartCheck as a current passenger option. The episode shows a different vulnerability: a useful deployment can lose its place when the surrounding operation changes. Buyers can copy the advance-enrollment idea, but they also need a plan for changing checkpoints, layouts and passenger flows.
The next suspicious face is on a work call
In March 2026, iProov introduced its Workforce Solution Suite, covering remote hiring, onboarding, shared-device access, sensitive approvals and account recovery. It is designed to work with existing identity and access systems. The commercial argument is that possession of a credential does not settle every question about the person using it.
May brought Verified Meetings. The product puts host-triggered checks inside video conferencing: imagery analysis looks for deepfakes and presentation attacks, while a separate check examines whether the feed originates from a physical camera. The host receives red, amber or green feedback. This gives the company another place to apply its expertise, before a candidate is hired or an important decision is made on a call.
That is a meaningful expansion of the job. A bank’s enrollment flow is a designed checkpoint; a meeting is a conversation. The value of an in-call signal depends on what people do with it. As an operational inference, organizations need to decide when a concerning result pauses an interview or sends an approval through another channel. A colored indicator alone does not write that procedure.
The face scan has a maintenance crew
iProov competes in a market where liveness is already a category. FaceTec, for example, offers 3D face matching and liveness built around its FaceMap approach. iProov’s distinguishing pitch combines controlled illumination with cloud analysis and its Security Operations Center, or iSOC. The latter monitors attack methods and supports updates to defenses.
The distinction matters because attackers can challenge both the image and its route into the system. Holding a photograph in front of a camera is different from injecting fabricated imagery into a video feed. Buyers should compare the attacks each product is tested against, the user experience and how improvements reach deployed systems. A generic promise to “stop deepfakes” is too broad to make that purchasing decision.
The working environment described by iProov has some concrete provisions behind the usual careers-page language. It lists coaching with an occupational psychologist, monthly knowledge-sharing sessions and a personal-development allowance of £250 for UK employees, with $325 for US staff. These are the company’s descriptions of its benefits. They fit a business where learning new attack techniques is part of maintaining the product.
Copy the checks around the check
For a team implementing biometric verification, one of iProov’s most useful instructions is plain: validate the result from the backend before granting privileges. Its Web SDK documentation explicitly warns against trusting the browser’s success message. A convincing front-end experience is not sufficient evidence for the server making the access decision.
The same documentation describes device-support checks and alternative journeys for unsupported users. Camera permissions, browser capabilities and network connectivity are practical conditions for success. Teams should test those conditions before sending people into a scan, and provide another route when they cannot complete it. An identity product that strands legitimate users creates a new service problem while solving an old security one.
The lesson running through iProov’s history is practical rather than mystical. Establish which fact a check proves. Keep that proof attached to the current session. Maintain the defenses and the recovery path. And remember the railway station: even when the face check works, the world around it can rearrange the queue.