Verification of stochastic gradient descent

This invention describes a method to ensure that updates to a machine learning model are transparent and verifiable. A data owner computes changes to the model using their private data, encrypts this update information, and then records it on a blockchain. This creates a secure, unchangeable record that can be audited to confirm the model's evolution without exposing the underlying private data. The claims narrow this to a computer system performing these specific steps.

Why it matters: In 2020, the practical demand for auditable and verifiable machine learning updates on a blockchain was still emerging. By 2026, the increased regulatory scrutiny on AI transparency and accountability, coupled with more mature enterprise blockchain solutions, has made such verifiable and privacy-preserving ML systems far more relevant.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedSoftware logic, blockchain, cryptography, machine learning

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10