QuantumLearning Machines
CAE evidence workspace

CyberSim · CAE-CAI 2026

Knowledge Unit practicum

Complete the actual work, attach its outputs, explain what the evidence shows, and transfer the method to an unfamiliar case. Faculty review every official outcome.

CSF · threat lab · 70 minutes

Attack-to-control range

Investigate a compromised AI support service, classify the attack path, rank vulnerabilities, and deploy proportionate controls.

Do the work

  1. Build an attack path from mixed host, identity, and model records.
  2. Rank three vulnerabilities by impact and exploitability.
  3. Configure access, integrity, and recovery controls, then rerun the incident.

Submit these artifacts

  • Attack-path graph with cited records.
  • Before-and-after control test report.

Server-executed native workbench

Run the primary and transfer fixtures

The server executes each fixture and returns a signed, content-addressed artifact that is sealed with the submission.
Primary fixturecae-practicum-csf
Unfamiliar transfercae-practicum-csf-transfer

Outcome evidence

Artifact references can be approved repository URLs, LMS files, or institution-controlled evidence identifiers.