Menu
Conscious of this reality, World wants to scale biometric identity globally, safeguarding user privacy and confidentiality by design.
World partnered with Inversed to understand whether privacy-enabled biometric identity solutions could be feasible at a truly global scale.
In a digital landscape increasingly influenced by automation and AI, identifying humans accurately becomes even more essential. World's vision is to provide a privacy-preserving identity solution for everyone. Yet, accurate biometric systems are complex to scale. Especially when prioritizing the preservation of user privacy while maintaining acceptable performance and cost.
World wanted to figure out whether a software solution is capable of maintaining robust privacy at a global scale. Enabling biometric verification without unsustainable expenditures. No shortcuts. No compromises.
Inversed Tech partnered with World to research, prototype, and validate MPC-based solutions for biometric verification. We leveraged Multi-Party Computation (MPC), to power biometric authentication that doesn’t reveal or centralize sensitive data. Building a scalable system that doesn’t rely on specialized hardware.
Implementing this MPC-driven approach allowed World to confidently pursue global biometric authentication. Balancing security, scalability, and privacy– to help realize their commitment to a universally accessible identity verification system.
World's ambitious agenda faces a core technical challenge: biometric signals naturally vary because they're captured from real-world sensors. Matching identities isn't just a simple comparison — it requires nontrivial similarity computations that must be privacy-preserving and able to massively scale.
To find a practical path forward, our first phase of work was an open research collaboration, exploring a broad range of technologies for both privacy and scalability. We evaluated methods like fuzzy commitment schemes, error-correcting codes, secure multi-party computation (MPC), fully homomorphic encryption (FHE), oblivious RAM (ORAM), clustering, locality-sensitive hashing, and graph-based similarity search. From this, we built a privacy framework that supports integrated protection of sensitive data at its core, while enabling a scalable, high-performance matching layer. We rigorously vetted the design with small-scale prototypes and delivered a detailed analysis and viability report to the World team.
Deployment of the design has followed a two-phase journey. First, we collaborated with contributors from several organizations to finalize the base-level privacy layer and support the implementation of production-grade server software. We built an early prototype FHE engine for secure iris code comparisons, and contributed to the theoretical design of the MPC protocol. This ultimately became the core of the privacy solution. Second, we engineered and implemented the scaling architecture — a system employing the HNSW graph similarity search algorithm, whose logarithmic performance enables efficient processing; even as the uniqueness set grows to global scale. The rollout and optimization of this scaling solution are actively ongoing.
Though still early in implementation, this collaboration demonstrates the feasibility of a MPC-based biometric identity system at a truly global scale. Establishing a standard for secure biometric verification without centralized risks. Ultimately, positioning World as a leader in secure, decentralized, hardware-independent identity verification.
The project highlights the feasibility of maintaining robust privacy in large-scale biometric systems. World now has a clear roadmap to scale identity solutions securely. For Inversed, it represents a critical step towards a future where security is an inherent design principle. Never compromised by the demands of scale.
Private biometrics are clearing a path to secure, accessible identity for everyone within view. Paving the foundations of a freer, more trusted digital world.
The Inversed team didn’t just prototype with us; they truly helped push the thinking to a new level. The framework that is emerging from the project is a meaningful step toward inclusive, decentralized identity.
