Multi-Level Security Lattice: Bell-LaPadula for Government Systems

Subject Clearance

Clearance level SECRET

Request

Target document level TOP SECRET

Choose an action and press Send Request. Rule enforcement runs on every request, manual or automated.

Automated Request Stream

Requests / second 0.0

Session Statistics

Total requests

Allowed

Simple-Security denials

*-Property denials

How it works

Government and critical-infrastructure systems classify every document into an ordered set of levels and every user into a matching clearance. The Bell-LaPadula model (1973, still the formal basis of most government MLS/MAC systems) enforces confidentiality with two rules, given subject clearance L(s) and object classification L(o):

Simple Security Property (ss-property), "no read up":
  read(s, o) allowed  ⇔  L(s) ≥ L(o)

Star Property (*-property), "no write down":
  write(s, o) allowed  ⇔  L(s) ≤ L(o)

Levels: 0 UNCLASSIFIED < 1 CONFIDENTIAL < 2 SECRET < 3 TOP SECRET

Reading is blocked upward so a low-clearance subject can never see high-classification data. Writing is blocked downward so a high-clearance subject can never leak sensitive content into a lower-classification document (deliberately or via malware) — together the two rules make information flow only "flow up or stay level," never down.

Clearance slider — sets the subject's clearance level (the sphere's shelf in the 3D lattice).
Target level slider — picks which classification tier's document the request targets.
READ / WRITE — selects which property is being tested, then Send Request fires an animated request pulse: green means the lattice allowed it, red means it was blocked, and the verdict box names the exact rule.
Automated stream — generates random subject/target/action requests at the chosen rate, the same way a SOC's policy-enforcement point evaluates a continuous flow of access attempts against the classification lattice.

Real systems layer more on top of this core (compartments/need-to-know via the Chinese Wall or lattice-of-labels model, audit trails, SIEM correlation of denied attempts) but the read/write inequality above is the mechanism every one of them ultimately checks.

Subject (clearance) Allowed request Denied request

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