Abstract
IEPP (Intrinsic Entropy Proof Protocol) introduces a structural method for existence verification based on intrinsic entropy and state continuity.
Unlike static identifiers, IEPP generates non-reproducible fingerprints through dynamic challenge–response interactions. Each response incorporates internal entropy, temporal state progression, and external challenge inputs to produce a continuously evolving cryptographic identity trace.
Experimental Proof-of-Concept (PoC) simulations conducted in reproducible Google Colab environments indicate stable uniqueness, replay resistance, and robustness across diverse entropy sources, including computational jitter environments.
IEPP aims to provide a foundational trust primitive for AI agents, digital entities, and distributed autonomous systems operating in environments where replication cost approaches zero.
1. Problem Statement
As generative AI systems scale, the marginal cost of producing synthetic content approaches zero. This leads to a structural authenticity imbalance, where distinguishing genuine outputs from replicas becomes increasingly difficult.
Traditional verification mechanisms assume static identity models. However, AI systems generate outputs dynamically, often without persistent physical anchors.
When replication becomes trivial, static identifiers lose meaning.
A verification method must therefore consider not only what an entity is, but whether the entity demonstrates continuity of existence.
IEPP explores a structural approach where existence is represented through evolving entropy-bound responses rather than fixed identifiers.
2. Entropy Fingerprint Concept
IEPP introduces the concept of an entropy fingerprint: a response pattern generated through interaction between intrinsic entropy, temporal state progression, and external challenge input.
Each response is unique due to the dynamic combination of:
- intrinsic entropy (E)
- external challenge (C)
- internal temporal progression (T)
- internal state continuity (S)
Experimental PoC simulations demonstrate 100% uniqueness across up to 5,000,000 iterations under varying entropy configurations, indicating strong non-reproducibility characteristics.
Unlike static hashes, entropy fingerprints evolve continuously, reflecting ongoing existence rather than static identity claims.
3. Entropy Challenge–Response Structure
IEPP responses are generated through a structured challenge–response interaction model.
Each response incorporates:
- internal entropy source
- external challenge input
- temporal variation
- state progression continuity
This creates a response trajectory that is difficult to replicate without access to the internal entropy and state continuity.
Observed PoC results indicate 0% successful replay under tested attacker models, supporting structural replay resistance characteristics.
4. Core Formulation
A simplified representation of the response structure is:
R = H(E || C || T || S)
Where:
E = intrinsic entropy component
C = external challenge
T = temporal state input
S = internal continuity state
H = cryptographic hash function
The formulation emphasizes structural non-repeatability rather than secrecy of algorithm.
Security derives from entropy availability and state continuity progression.
5. Experimental Proof-of-Concept Validation
A multi-stage Proof-of-Concept evaluation was conducted using reproducible Google Colab environments.
Test objectives
- verify uniqueness properties
- evaluate replay resistance
- assess entropy source independence
- analyze multi-prover scalability
- simulate adversarial response attempts
Iteration scale
| test | iterations | observed result |
|---|---|---|
| uniqueness test | 1,000,000 | 0 collisions |
| stress test | 5,000,000 | stable |
| multi-prover simulation | 200,000 requests | 100% verified |
Processing throughput observed:
up to approximately 35,000 operations per second in Python runtime environments.
These results indicate stable structural uniqueness under repeated evaluation conditions.
Deterministic control validation
Under fully controlled conditions:
- fixed entropy seed
- fixed challenge
- fixed time
- fixed state
the system produced identical fingerprints across repeated trials.
When state progression or temporal variation was introduced, fingerprints diverged consistently.
This confirms both:
controlled reproducibility
state-driven non-repeatability
Entropy source robustness
The following entropy configurations were evaluated:
- operating system randomness
- pseudorandom generators
- numpy random
- torch random
- mixed entropy sources
Observed result:
consistent 100% uniqueness across tested entropy source variations.
This indicates that the structural behavior of IEPP is not dependent on a single randomness implementation.
Multi-prover simulation
Simulations involving 50 independent provers produced:
- independent state progression
- zero observed fingerprint collisions
- stable verification behavior under shared challenge conditions
Memory usage remained stable during multi-entity testing.
These results suggest scalability characteristics compatible with distributed identity environments.
Attacker model simulation
Simulated attacker assumptions:
attacker knows:
- algorithm structure
- challenge input
- prior response outputs
attacker does not know:
- internal entropy values
- internal continuity state
- secret binding values
Tested attack categories:
- replay attack
- prior response reuse
- challenge substitution attempt
- deterministic imitation attempt
- guessed secret attempt
Observed result:
0% successful forgery across tested attacker models.
Primary rejection causes:
replay detection
fingerprint mismatch
non-monotonic state progression
These observations indicate structural resistance to basic forgery strategies.
Physical-like entropy simulation
Physical-like entropy was simulated using:
- timing jitter
- floating point computation variation
- latency fluctuation
- mixed computational perturbations
Observed results:
100% fingerprint uniqueness
consistent entropy variability across runs
These results suggest that computational environments may provide usable entropy signals for dynamic fingerprint generation.
6. Security Characteristics
Observed structural properties include:
non-reproducible response generation
replay resistance under tested attacker assumptions
state continuity sensitivity
entropy source flexibility
multi-entity scalability
IEPP does not rely on secrecy of algorithm structure.
Security derives from entropy availability and continuity state progression.
7. Positioning relative to existing approaches
| property | IEPP | digital signature | hash |
|---|---|---|---|
| replay resistance | yes | partial | no |
| state continuity | yes | limited | no |
| entropy binding | yes | no | no |
| AI-native identity | yes | limited | no |
| dynamic existence trace | yes | no | no |
IEPP operates between traditional challenge-response authentication and physical unclonable function concepts, extending applicability toward AI-native identity environments.
8. Implementation status
Core structural concepts have been filed under PCT framework.
Additional implementation details may be disclosed progressively following filing milestones.
Current results represent experimental validation under controlled environments.
Further real-world evaluation is ongoing.
9. Scope
IEPP is not proposed as a replacement for existing cryptographic primitives.
Instead, IEPP explores an additional structural layer enabling dynamic existence verification where replication cost approaches zero.
Potential application domains include:
AI agents
digital entities
distributed autonomous systems
device identity environments
virtual avatars
synthetic content provenance
Closing note
When replication becomes trivial, continuity becomes meaningful.
IEPP explores whether continuity itself can function as a verifiable signal.