| CPC G06F 21/6227 (2013.01) [G06F 21/604 (2013.01); G06F 2221/2113 (2013.01)] | 20 Claims |

|
1. A computer-implemented method, comprising:
inputting data intended for storage into a trained artificial intelligence (AI) model to derive data access privilege grants for the data before the data is stored in predetermined storages, wherein user information for training the trained AI model is selected from among profile information of a given user, credentials held by the given user, usernames of the given user, strength of a password used by the given user, experience with a predetermined subject of the given user, demographic information of the given user, personally identifiable information of the given user, and financial information of the given user;
in response to deriving the data access privilege grants for the data and prior to storing the data, segmenting the data to correspond to the data access privilege grants as segments of the data are being stored according to the data access privilege grants in the predetermined storages, wherein metadata that defines the data access privilege grants is stored with the segments of the data;
receiving, from a first user device, a query requesting the data stored in the predetermined storages;
reading at least some of the metadata to identify a segment of the data associated with a data access privilege grant associated with the query; and
allowing the first user device to access the identified segment of data for fulfilling the query.
|