Skip to content

Privacy Finality Reference v0.1.0

Latest

Choose a tag to compare

@sangmdas sangmdas released this 05 Sep 09:16
ecde09e

privacy-finality-reference-v0.1.0.tar.gz

Privacy Finality Reference v0.1.0 is a runnable Python reference implementation of Candidate-Act-based, purpose-bound privacy and data-use enforcement for AI systems.

This release demonstrates:

  • Candidate Act exact-act binding
  • Non-Effective State
  • logical non-joinable vaults
  • purpose and minimum-data enforcement
  • protected Virtual Identity commitment
  • non-bearer proof-of-possession
  • policy epoch and nonce checks
  • replay protection
  • Finality Sink verification
  • fail-closed release behavior
  • SME customer-support example
  • automated tests and microbenchmark

This is a research and interoperability reference implementation. It is not a production security product, legal-compliance engine, or grant of patent implementation rights.