USPTO Patent Grant: Minimal Trust Data Sharing

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published March 24th, 2026

Detected March 25th, 2026

Summary

The USPTO has granted a patent (US12585955B2) to Pulselight Holdings, Inc. for a computer-implemented method of protecting confidentiality during the generation of synthetic training records for machine learning using a distributed generative adversarial network (GAN). This method aims to generate synthetic data without revealing the contents of sensitive source data records.

View original document View source feed page

What changed

The United States Patent and Trademark Office (USPTO) has issued patent US12585955B2 for a method of minimal trust data sharing, assigned to Pulselight Holdings, Inc. The patent covers a computer-implemented process for generating synthetic training records for machine learning models from sensitive data. The core innovation involves distributing the generator and discriminator components of a generative adversarial network (GAN) across separate source and target computer systems, ensuring that the sensitive data records on the target system are not revealed during the synthetic data generation process.

This patent grant signifies a technological advancement in data privacy for AI development. While it does not impose new regulatory obligations on companies, it provides a patented method that may influence how organizations approach synthetic data generation for machine learning, particularly in highly regulated sectors where data confidentiality is paramount. Companies developing or utilizing similar technologies should be aware of this granted patent.

Source document (simplified)

← USPTO Patent Grants

Minimal trust data sharing

Grant US12585955B2 Kind: B2 Mar 24, 2026

Assignee

Pulselight Holdings, Inc.

Inventors

Jonathan Mugan, Mallika Thanky

Abstract

A computer-implemented method of protecting confidentiality when generating synthetic training records for machine learning from sensitive data records, comprising a source computer system connected to a separate target computer system, where the target computer system comprises sensitive data records comprising private or confidential data. The source computer system performs the functions of a generator component of a generative adversarial network (GAN) and the target computer system performs the functions of a discriminator component of the GAN, where the generator and discriminator functions of the GAN are distributed between the source and target computer systems. Synthetic training records are generated using a computational process that does not reveal contents of the sensitive data records. Also disclosed is a method of training a machine-learning model using the one or more synthetic training records.