US 12,469,027 B2
Artificial intelligence model and dataset security for transactions
Rakesh Ramde, Los Altos, CA (US); and Amanjyot Singh Johar, Sunnyvale, CA (US)
Assigned to Datacurve, Inc., Los Altos, CA (US)
Filed by Datacurve, Inc., Los Altos, CA (US)
Filed on May 20, 2024, as Appl. No. 18/669,109.
Claims priority of provisional application 63/467,711, filed on May 19, 2023.
Prior Publication US 2024/0420124 A1, Dec. 19, 2024
Int. Cl. G06Q 30/0645 (2023.01); G06F 21/62 (2013.01); G06Q 20/36 (2012.01); G06Q 20/38 (2012.01)
CPC G06Q 20/38215 (2013.01) [G06F 21/6245 (2013.01); G06Q 20/3674 (2013.01); G06Q 30/0645 (2013.01); G06Q 2220/00 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method in a system, on a data communication network, for providing immutable information about artificial intelligence (AI) models and their training datasets, deployed by an entity, secured and observed with a non-fungible cryptographic token (NFT) based observation certificate, the method comprising:
creating an entity private key/wallet address pair associated with a specific enterprise;
creating a first smart contract using the entity private key/wallet address pair;
using the first smart contract, generating a backpack NFT token representing the AI model;
creating a second smart contract that binds to the backpack NFT token;
using the first smart contract, generating a dataset NFT token in a chronological series of tokens, representing the training dataset that the AI model was trained with;
wherein the dataset NFT token is bound to the second smart contract and the backpack NFT token; and
taking an action, responsive to the backpack NFT token and the training dataset NFT token of the series of tokens.