Security Program

Continuous adversarial testing against state-level synthesis attacks. Methodology public, defenses layered, results measured.

// THREAT MODEL

Who we build against

We assume a well-resourced adversary with access to modern voice cloning (XTTS-v2, F5-TTS, ElevenLabs), generative models for biometric time-series, full source-code access to our public components (SDK, circuits, on-chain programs), unlimited wallets and devnet SOL, and days to weeks of time per attack campaign.

We do not assume the adversary can compromise user devices, mount physical hardware attacks on phones, or access our private defense-layer infrastructure. Those are separate threat categories covered by standard client-side hardening, hardware root-of-trust guidance, and infrastructure security practice respectively.

// DEFENSE LAYERS

How we defend

Tier 0Cryptographic gate

Zero-knowledge proofs of behavioral consistency. Groth16 proving system. Public verifier on Solana. Every verification produces a proof that the user’s behavioral fingerprint is within a hidden Hamming distance threshold of their baseline, without revealing either fingerprint. Open source, auditable, verifiable on-chain.

Tier 1Statistical distribution checks

Server-side validation of the 134-dimensional feature vector extracted from each verification. Multiple independent checks verify that the statistical properties of extracted features are consistent with human physiology, not synthetic generation. Specific checks and threshold values are not published.

Tier 2Behavioral coupling signals

Time-series analysis of phonation and kinematic signals sampled during capture. Real human speech and hand motion share motor-cortex origins and produce measurable temporal coupling at short lags; independent synthesis does not. Enforcement is live on production since April 2026, calibrated against a two-wave red team study that isolated the layer's specific contribution to voice-replay rejection.

// RED TEAM

How we test our defenses

We maintain an internal adversarial testing harness that runs continuously against our production verification pipeline. The harness implements eight distinct attack tiers, ordered by sophistication. Each campaign generates hundreds to thousands of bot attempts, measures pass rates per defense layer, and feeds the results into threshold calibration and defense roadmap decisions.

TierAttack classTests defense against
T1Procedural synthesis (script-kiddie baseline)Absolute attacker floor
T2Parameter-varied proceduralTier 1 statistical consistency checks
T3Feature-space optimization with source accessTier 1 distributional realism
T4aPre-recorded human voice + procedural motion/touchCross-modal temporal coupling (Tier 2)
T4bModern voice cloning (XTTS-v2, F5-TTS, API-based)Tier 1 TTS artifact detection
T5Coupled cross-modal synthesisTier 2 temporal coupling
T6Targeted human-mimicry / identity theftHamming distance gate + Sybil registry
T7Replay with adversarial perturbationMin-distance floor + commitment registry
T8Black-box adaptive probingRate limits + response opacity

Attack implementation code, per-attempt telemetry, and specific parameter values that produce elevated pass rates are kept in a private repository. This follows the same disclosure convention used by infrastructure security programs: methodology public, weapons private.

// MEASUREMENTS

Current measurements

Attack tierDescriptionAttemptsPass rateStatus
T1Procedural synthesis2,0000%hardened · 2026-03
T2Multi-strategy parameter variation4,0000%hardened · 2026-03
T3aUnconstrained feature optimization1,0000%hardened · 2026-04
T3bConstrained feature optimization9,0000%hardened · 2026-04
Campaign surfaced a gap in server-side feature validation. Hardened — see AUDIT.md.
T4a — Wave 1Pre-recorded human voice + procedural motion/touch (temporal enforcement OFF — log-only)50100%counterfactual baseline
T4a — Wave 2Pre-recorded human voice + procedural motion/touch (temporal enforcement ON)1010%production enforcement truth
Cross-program binding gap in update_anchor discovered during cross-analysis, patched same day — see AUDIT.md protocol-core Critical.
T4bModern voice cloning (XTTS-v2, F5-TTS)queuednext-phase
T5Coupled cross-modal synthesisqueuednext-phase
T6Targeted human-mimicry / identity theftqueuednext-phase
T7Replay-perturbedqueuednext-phase
T8Adaptive probingqueuedpost-mainnet

Last updated: April 20, 2026

Pass rate = fraction of bot attempts that pass server-side Tier 1 validation, the gate preceding on-chain submission. An attempt that fails Tier 1 cannot proceed to challenge fetch, proof generation, or transaction submission. Results rounded to prevent adversarial threshold inference.

// T4A — TWO-WAVE STUDY

T4a was designed as a two-wave study to measure the cross-modal temporal coupling layer's specific contribution to the multi-layer defense. Wave 1 ran with temporal enforcement in log-only mode to establish the counterfactual baseline. Wave 2 ran with enforcement enabled. The 90 percentage-point reduction isolates that layer's contribution. The 10% Wave 2 residual motivated server-side phrase content binding (shipped 2026-04-25) as the next defense layer; T4a Wave 3 is queued to measure the residual after that layer.

// ON-CHAIN ANCHOR STATE

The Entros Anchors currently visible on devnet include internal red team artifacts from T4a Waves 1–2 (documented above) alongside legitimate team and pilot-user verifications. All state is preserved on-chain for audit traceability; the public /stats page reads the full on-chain aggregate directly.

// OPEN SOURCE

What we open-source, and why

Entros Protocol is open-source where open-source matters for user trust, and deliberately private where privacy protects users. This follows the same disclosure convention used across crypto infrastructure projects. Not a departure from crypto's open-source values. A mature implementation of them.

Open source (MIT)

  • On-chain programs (entros-anchor, entros-verifier, entros-registry)
  • ZK circuits and trusted setup artifacts
  • Client SDK (pulse-sdk on npm)
  • Executor node
  • Website and documentation
  • Security program page, blueprint documents, and aggregate results
  • Baseline adversarial testing (script-kiddie tier in pulse-sdk)

Private (defense-layer only)

  • Server-side validation service (entros-validation): check thresholds and parameter values
  • Red-team harness (entros-redteam): attack code, per-attempt telemetry, captured baseline fixtures
  • Pre-disclosure vulnerability reports (per standard responsible-disclosure practice)

Nothing that affects verifiable protocol behavior is private. Every on-chain transition, every cryptographic operation, every client-side computation is open and auditable. The private components are the detection surface an attacker would otherwise exploit to calibrate their attacks.

// DISCLOSURE

Reporting vulnerabilities

ScopeOn-chain programs, SDK, executor, validation service, website
Response SLAAcknowledgment within 48 hours, initial triage within 5 business days
Safe harborGood-faith research is welcome. We will not pursue legal action against researchers acting within the scope of this policy.
AttributionResearchers credited in AUDIT.md and hall of fame upon fix deployment, unless anonymity requested.
Bug bountyPlanned post-launch. Severity tiers and amounts TBD.
// REFERENCES

Learn more


Read the research behind our verification pipeline.