On how it works, security, pricing, and getting started.
Sealfie stops deepfake fraud and executive impersonation by verifying the real requester's identity with a selfie. When someone requests a wire transfer or sensitive action, Sealfie confirms it's really that person — not a criminal using a deepfake or stolen credentials. In 2024, ARUP lost $25M after a video call where every other participant was a deepfake.
When you receive a sensitive request (wire transfer, credential change, decision), you ask the person to verify with Sealfie. They take a selfie, and Sealfie confirms their identity while checking, in parallel, the mobile carrier (SIM-swap detection), device fingerprint and the physical presence of the real person. The result is sealed into a timestamped log.
Your biometrics never leave your phone: the fingerprint or face recognition is processed by your operating system, inside the device's secure element (Secure Enclave on iPhone, StrongBox/TEE on Android). It only authorises the use of a cryptographic key that stays confined to that element. We never receive, process or store any biometric data. The selfie is kept as a timestamped visual record, with no template extracted and no facial recognition. Subprocessors are detailed in our privacy policy.
Sealfie costs €95 excl. VAT per user per month, with 20 verifications a year (most CFOs use 4 to 6). A custom-quote Enterprise plan exists for higher volumes, multiple approval chains and Attesta (video-call verification).
Book the free 30-minute Challenge: we analyze your current procedure, show how it would be bypassed, and give you a concrete plan — whether or not you choose Sealfie.
Contact support to suspend the verification tied to your device. Recovery is handled as a priority by the team.
No — only executives and high-risk people (CFO, treasury, decision-makers) need Sealfie. Sizing depends on your actual org chart, not a generic package.
No, it complements it differently: training aims for humans to spot fraud, Sealfie verifies automatically without depending on human judgment — which matters against deepfakes designed to fool trained professionals.
Because even after training, people identify only 73% of speech deepfakes — more than a quarter get through, and training improves results only marginally (UCL, PLOS ONE, August 2023). Tools fare no better: a CSIRO / Sungkyunkwan 2025 study benchmarked 16 reference detectors and none reliably identified real-world deepfakes. Sealfie doesn't ask anyone to spot anything by eye: verification is automatic, before the question even arises.
Standard MFA relies on methods that can be bypassed: SMS codes intercepted via SIM-swap, push notifications approved through fatigue, or known flaws like AuthQuake (a Microsoft MFA bypass in about an hour with no user interaction — fixed by Microsoft in October 2024). Sealfie cross-checks several independent sources at every verification — biometrics, mobile carrier, device fingerprint — instead of relying on a single factor.
Sealfie