Skeys !new! -

Abstract Traditional cryptographic secrets (keys, passwords, tokens) are static, stored separately from the user, and vulnerable to theft, loss, or coercion. This paper introduces Skeys (Somatic + Situational Keys)—a novel class of ephemeral, context-dependent secrets dynamically generated from a fusion of real-time behavioral biometrics, environmental data, and user intention modeling. Unlike conventional keys, Skeys exist only during a user’s interaction with a system and dissolve upon context change (e.g., location shift, physiological state change, or task termination). We present a mathematical model for Skey generation using recurrent neural networks over continuous sensor streams, propose a zero-knowledge proof protocol for Skey-based authentication without secret reconstruction, and analyze attack surfaces including adversarial sensor spoofing. We conclude with a proof-of-concept implementation using smartwatch PPG (heart rate), gaze tracking, and typing cadence, achieving an equal error rate of 0.03% in continuous authentication scenarios. 1. Introduction Current cryptographic systems suffer from a secret-storage paradox : secrets must be long-lived enough to be useful, yet short-lived enough to limit exposure. Passphrases are reused; hardware tokens can be stolen; biometrics (fingerprints, iris) are immutable once leaked. Skeys resolve this paradox by binding secrets to the continuous, non-repeating stream of user behavior and environment —what we call the context signature . The paper explores: (1) How can a secret be a process rather than an object? (2) Can authentication be both stateless and non-replayable? (3) What happens when we treat key generation as an emergent property of human-in-the-loop activity? 2. Definition & Properties A Skey is defined as:

Here’s a concept for a speculative, interdisciplinary paper titled: We present a mathematical model for Skey generation