Inkan.link is relevant to the AI agent ecosystem because it addresses the most significant risk of autonomous systems: unauthorized execution. As agents gain the ability to move money and interact with ERP systems, the need for a "human-in-the-loop" verification that cannot be spoofed by other AI becomes paramount. Sealfie provides a mechanism for humans to issue a biometric "proof of intent" that can authorize or throttle agentic actions.
In the agent stack, Inkan.link sits at the security and verification layer. They are championing the idea that as AI agents become more indistinguishable from humans in digital communication, the only reliable way to verify a request is to move outside the digital-only channel and require a physical biometric signal. For builders of autonomous finance agents, integrating such a verification layer is likely to become a requirement for enterprise adoption.
Inkan.link is a French cybersecurity startup that operates at the intersection of biometric identity and financial security. Founded in 2021 and headquartered in Crolles, the company targets a specific, high-cost vulnerability: the "Fake President" or Business Email Compromise (BEC) scam. As generative AI makes it trivial to clone an executive's voice or create realistic video deepfakes, traditional security measures like email filtering or even voice confirmation are becoming insufficient. Inkan.link's core product, Sealfie, is designed to close this gap by requiring a physical, biometric anchor for high-stakes financial transactions.
Sealfie works by introducing a multi-source authentication step into the payment workflow. When a finance professional receives a payment request—especially one that appears urgent or high-value—they use the Sealfie app to verify the identity of the requester. This involves a biometric check that takes roughly 30 seconds. The system is designed to be used in parallel with existing ERP systems, ensuring that the usual data entry and processing continue while the identity verification happens in the background.
By tying the validation to the physical identity of the validator, the company claims it can reduce manual validation time by 80%, moving from hours of manual phone calls and cross-referencing to a standardized digital process. The technology is specifically tuned to detect advanced threats, including voice deepfakes and AI-generated impersonation that can bypass legacy security layers.
The company gained significant international visibility with its debut at CES 2025. This launch occurred against a backdrop of increasing financial losses due to BEC, which the company cites as reaching $5 billion annually. While many identity verification startups focus on the onboarding phase (KYC), Inkan.link focuses on the operational phase—securing the actions taken by already-vetted employees.
Inkan.link is one of several startups emerging from the European tech ecosystem that focus on "liveness detection" and biometric verification. Their specific focus on the finance department's workflow distinguishes them from broader identity providers like Okta or Duo, which manage access but often lack the deep integration into specific transactional proofs of intent.
The competitive landscape for Inkan.link includes traditional cybersecurity firms and newer AI-safety startups. However, their primary competition is often the status quo: manual verification. CFOs currently rely on "four-eyes" principles and manual callbacks to verify suspicious requests. Inkan.link argues that these manual processes are both slower and more prone to social engineering than a biometric-first digital approach. By providing a 30-second validation window, they aim to lower the friction of security, making it more likely that employees will actually follow safety protocols rather than circumventing them for the sake of efficiency.
A payment authentication platform that uses biometrics to prevent deepfake-driven executive impersonation fraud.
Inkan.link is hiring