Video summary

Keyless Presentation at Gartner IAM Summit 2021

Main summary

Key takeaways

Technology

Summary (tech concepts, product features, analysis)

Problem & trend focus (identity / strong authentication)

  • Enterprises are rapidly moving toward passwordless authentication because passwords and weak credentials account for most breaches (noted as >80% from lost/stolen/weak credentials).
  • A central tension remains the trade-off between:
    • User experience (convenience)
    • Security and privacy, including reducing account takeover, phishing/fraud
    • Support for increasingly distributed work and heterogeneous devices

Proposed direction: “zero knowledge biometrics” & user-controlled identity

  • Keyless proposes a privacy-preserving approach called zero knowledge biometrics.
  • The goal is identity verification across any device/touchpoint via a simple camera-based look.
  • The system aims to do so without processing or storing biometric templates/PII centrally.
  • It also points toward a broader evolution of identity:
    • Decentralized, user-centric identity management
    • Selective disclosure using cryptography and public-key infrastructure concepts

Why current biometrics are limited (local vs. centralized)

  • Local/on-device biometrics (e.g., secure enclave storage) are:
    • Hard to use universally across devices
    • More like “unlocking a device/token” than identifying the person globally
    • Potentially insufficient for requirements around universal identity compliance and security
  • Central biometric matching can improve universality, but introduces:
    • Privacy and compliance concerns, since it requires storing biometric data in a central database

Keyless approach: privacy-enhancing cryptography + biometric signals

Keyless combines multiple techniques:

  • Zero-knowledge proofs to prove device trust
  • Neural networks / on-device ML to transform biometric signals into irreversibly encrypted opaque objects
  • Secure multi-party computation (SMPC) and distributed cryptographic protocols to derive a usable credential

Outcome: it deterministically reconstructs a private key for one-time actions such as:

  • Logging into an enterprise app
  • Connecting to SSO

Crucially, it claims this avoids storing biometric data and secret key material:

  • on the device
  • or in centralized infrastructure

“Omnichannel biometrics” / multi-factor without extra hardware

  • Authentication is described as omnichannel, working across devices/platforms (independent of OS/hardware).
  • It uses passive facial recognition with strong liveness detection, explicitly claiming resistance to video/picture spoofing.
  • It adds behavioral characteristics (how the user interacts with the device) as an additional factor.

Multi-factor composition:

  • Factor of possession: trusted device identity proven via zero-knowledge proof (device secret; not user biometrics)
  • Inherence/identity: live biometric + behavioral pattern captured via the camera
  • No passwords remembered and no extra token required

Product/solution scope (workforce + consumer)

Workforce authentication suite (employees)

  • Passwordless MFA to SSO
  • Desktop MFA for Windows/workstations using:
    • a companion device/push notification, or
    • the built-in laptop camera (no additional hardware)
  • Remote login for:
    • VPNs
    • remote desktops

Consumer / regulated use cases

  • Delivered as an SDK
  • Focus on PSD2 / Open Banking “Strong Customer Authentication”
  • Enables frictionless onboarding + ongoing MFA, with adjacent cases such as digital signatures

Claims on compliance & privacy

  • Strong emphasis that the system:
    • does not process personal identifiers
    • involves no PII in the authentication pipeline
  • Claims it is compliant and “far exceeding” regulations like GDPR because it removes the central burden of storing sensitive biometric/PII data.

Deployment & partner integrations (examples & reported speed)

  • Education: Lewis University + Cisco
    • Remote exam proctoring/authentication from students’ homes
    • Reported as higher assurance than local device biometrics
  • Banking / VPN: European bank
    • Remote workforce VPN authentication using standard RADIUS integration
    • Deployed in hours; live within a day for ~1,000 employees
  • Telco / cloud edge acceleration: work with global telcos + Amazon
    • Claims to reduce authentication latency/download delays to milliseconds
    • Uses mobile edge computing / 5G-enabled architecture with distributed, decentralized computation

Overall market framing & “personal identity management”

  • Frames the technology as biometrics-enabled key management and personal identity management using:
    • zero-knowledge proofs
    • selective disclosure
  • Positions the shift from managing many passwords to managing keys tied to the individual as a major opportunity/challenge the approach is designed to solve.

Main speakers / sources

  • Speaker: Fabian — Co-founder of Keyless (presenting at/for Gartner IAM Summit 2021)

Original video