Zero Trust: Continuous Trust Scoring Gate

Verification Model

Every instant, the Policy Decision Point re-scores each session's trust. A session admitted through the gate can still be revoked mid-flight if its live score decays below the threshold.

Policy Parameters

Trust threshold τ 0.55

Trust decay rate λ 0.35 /s

Anomalous-behavior rate 25%

Request rate 6 / s

Live Readouts

Admitted → resource

Blocked / revoked

Avg. live trust score (in-flight)

0.00

How it works

NIST SP 800-207 defines Zero Trust Architecture around one core tenet: no implicit trust is ever granted based on network location. Instead a Policy Decision Point (PDP) computes a trust score for every access request from live signals, and a Policy Enforcement Point (PEP) — the gate you see around the resource — allows or denies traffic per request, continuously, not just once at the network edge.

trust(t) = [ wd·health + wl·(1 − locationRisk) + wb·(1 − anomaly(t)) ] · e^(−λt)

access granted  ⇔  trust(t) ≥ τ   — evaluated at EVERY instant t
wd=0.40  wl=0.35  wb=0.25   (device posture, network context, behavior)

Each simulated request carries a device-health and location-risk value drawn when it spawns, plus a chance of an anomalous-behavior event partway through its flight. The exponential term e^(−λt) models session-age decay: a credential or context established further in the past is inherently less trustworthy, so trust erodes even with no new evidence.

Zero Trust mode — trust(t) is re-evaluated continuously. A request only reaches the resource while trust(t) ≥ τ at every moment; a spike in anomalous behavior or enough decay after admission triggers live revocation (shown bouncing back, turning red).
Perimeter mode — trust is checked once, at t=0, before the request even starts its flight. If it passes that single check, nothing after the gate can stop it — even if the same anomaly later appears, exactly like a VPN/firewall perimeter that trusts anyone already inside.
Trust threshold τ — the PDP's minimum acceptable score; raising it enforces stricter least-privilege access.
Decay rate λ and anomaly rate — how aggressively context "goes stale" and how often a session exhibits risky behavior mid-session.

Real-world relevance: this is the same request-time evaluation loop used by production Zero Trust systems (Google BeyondCorp, Microsoft Conditional Access, service-mesh mTLS policy engines) — the score, weights and thresholds differ, but "never trust, always verify, at every request" is the shared mechanism.

High trust / admitted Mid trust / in-flight Low trust / blocked-revoked

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.