# US 2026/0064890 A1 · Training Data Provenance System and Method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0064890 A1
- Document ID: US20260064890A1
- Status: Published application
- Kind: A1
- Application number: 18/905,914
- Filed: 2024-10-03
- Published: 2026-03-05
- Earliest claimed: 2024-09-03
- Assignees: Microsoft Technology Licensing, LLC
- Applicants: Microsoft Technology Licensing, LLC
- Inventors: Mark Russinovich; Bryan D. Kelly
- Web page: https://patentgrep.com/patent/US20260064890A1

## CPC

- G06F 21/64: Protecting data integrity, e.g. using checksums, certificates or signatures
- G06F 21/602: {Providing cryptographic facilities or services}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A method, computer program product, and computing system for generating signed data for training an artificial intelligence (AI) model by processing data stored on a ledger using a signing authority. Signed firmware is generated for training the AI model by processing data stored on the ledger using the signing authority. The AI model is trained with signed data and the signed firmware from the ledger using a data processing unit in response to determining that the signed data and the signed firmware are signed by the signing authority.

## Claims

### Claim 1

A computer-implemented method, executed on a computing device, comprising:
generating signed data for training an artificial intelligence (AI) model by processing data stored on a ledger using a signing authority;
generating signed firmware for training the AI model by processing data stored on the ledger using the signing authority; and
training the AI model with signed data and the signed firmware from the ledger using a data processing unit in response to determining that the signed data and the signed firmware are signed by the signing authority.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein training the AI model includes preventing the training of the AI model using the data and the firmware in response to determining that at least one of the data and the firmware are unsigned by the signing authority.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1, wherein the data processing unit is a graphics processing unit (GPU).

### Claim 4 (depends on 1)

The computer-implemented method of claim 1, further comprising:
verifying the data stored on the ledger prior to generating the signed data using the signing authority.

### Claim 5 (depends on 1)

The computer-implemented method of claim 1, further comprising:
generating a signed data processing unit identifier for the data processing unit by processing a data processing unit identifier stored on the ledger using the signing authority.

### Claim 6 (depends on 5)

The computer-implemented method of claim 5, wherein training the AI model includes training the AI model with the signed data and the signed firmware from the ledger using the data processing unit in response to determining that the signed data processing unit identifier is signed by the signing authority.

### Claim 7 (depends on 1)

The computer-implemented method of claim 1, wherein the signing authority uses public private key cryptography for generating the signed data and the signed firmware.

### Claim 8

A computer program product residing on a non-transitory computer readable medium having a plurality of instructions stored thereon which, when executed by a processor, cause the processor to perform operations comprising:
verifying data stored on a ledger for training an artificial intelligence (AI) model;
generating signed data for training the AI model by processing the data stored on a ledger using a signing authority;
generating signed firmware for training the AI model by processing data stored on the ledger using the signing authority; and
training the AI model with signed data and the signed firmware from the ledger using a data processing unit in response to determining that the signed data and the signed firmware are signed by the signing authority.

### Claim 9 (depends on 8)

The computer program product of claim 8, wherein training the AI model includes preventing the training of the AI model using the data and the firmware in response to determining that at least one of the data and the firmware are unsigned by the signing authority.

### Claim 10 (depends on 8)

The computer program product of claim 8, wherein the data processing unit is a graphical processing unit (GPU).

### Claim 11 (depends on 8)

The computer program product of claim 8, wherein the operations further comprise:
generating a signed data processing unit identifier for the data processing unit by processing a data processing unit identifier stored on the ledger using the signing authority.

### Claim 12 (depends on 11)

The computer program product of claim 11, wherein training the AI model includes training the AI model with the signed data and the signed firmware from the ledger using the data processing unit in response to determining that the signed data processing unit identifier is signed by the signing authority.

### Claim 13 (depends on 8)

The computer program product of claim 8, wherein the signing authority uses public private key cryptography for generating the signed data and the signed firmware.

### Claim 14 (depends on 13)

The computer program product of claim 13, wherein the signing authority includes a private key and the data processing unit includes a corresponding public key.

### Claim 15

A computing system comprising:
a memory; and
a processor configured to:
generate signed data for training an artificial intelligence (AI) model by processing data stored on a ledger using a signing authority;
generate signed firmware for training the AI model by processing data stored on the ledger using the signing authority; and
train the AI model with signed data and the signed firmware from the ledger using a graphical processing unit (GPU) in response to determining that the signed data and the signed firmware are signed by the signing authority.

### Claim 16 (depends on 15)

The computing system of claim 15, wherein training the AI model includes preventing the training of the AI model using the data and the firmware in response to determining that at least one of the data and the firmware are unsigned by the signing authority.

### Claim 17 (depends on 15)

The computing system of claim 15, further comprising:
verifying the data stored on the ledger prior to generating the signed data using the signing authority.

### Claim 18 (depends on 15)

The computing system of claim 15, further comprising:
generating a signed data processing unit identifier for the data processing unit by processing a data processing unit identifier stored on the ledger using the signing authority.

### Claim 19 (depends on 18)

The computing system of claim 18, wherein training the AI model includes training the AI model with the signed data and the signed firmware from the ledger using the data processing unit in response to determining that the signed data processing unit identifier is signed by the signing authority.

### Claim 20 (depends on 15)

The computing system of claim 15, wherein the signing authority uses public private key cryptography for generating the signed data and the signed firmware.
