# US 12,008,134 B2 · Artificial intelligence-based data security management > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,008,134 B2 - Document ID: US12008134B2 - Status: Granted - Kind: B2 - Application number: 17/332,467 - Filed: 2021-05-27 - Published: 2022-12-01 - Granted: 2024-06-11 - Earliest claimed: 2021-05-27 - Assignees: Dell Products L.P. - Applicants: Dell Products L.P. - Inventors: Shibi Panikkar; Thiru Shama; Chetan Pudiyanda Somaiah - Web page: https://patentgrep.com/patent/US12008134B2 ## CPC - G06F 21/552: {involving long-term monitoring or reporting} - G06F 21/6245: {Protecting personal data, e.g. for financial or medical purposes} - G06F 21/31: User authentication - G06F 21/604: {Tools and structures for managing or administering access control systems} - G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] - G06N 20/00: Machine learning - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Artificial intelligence-based data security management techniques are disclosed. For example, a method comprises the following steps. A dataset is downloaded from a data source. The method detects whether or not the dataset contains at least a subset of data of a given security-based data type, wherein the detection is performed in accordance with an artificial intelligence model. A machine-computed security level classification is generated for the dataset by automatically selecting one of a plurality of security level classifications based on the detection. The machine-computed security level classification is compared with a user-generated security level classification manually selected from a plurality of security level classifications. One or more actions are caused to be taken based on the comparison. ## Claims ### Claim 1 An apparatus comprising: at least one processing device comprising a processor coupled to a memory, the at least one processing device, when executing program code, is configured to: download a dataset from a data source; detect whether or not the dataset contains at least a subset of data of a given security-based data type, wherein the detection is performed in accordance with an artificial intelligence model; generate a machine-computed security level classification for the dataset by automatically selecting one of a plurality of security level classifications based on the detection; compare the machine-computed security level classification with a user-generated security level classification manually selected from the plurality of security level classifications; cause one or more actions to be taken based on the comparison; and create a unique identifier for the dataset to track a location of the dataset, the unique identifier comprising a user identification and a timestamp; wherein, when the generated machine-computed security level classification is a restricted security level classification, the one or more actions comprise relocating the dataset based on the unique identifier from a device that downloaded the dataset to a secure storage location. ### Claim 2 (depends on 1) The apparatus of claim 1, wherein, when the comparison results in the machine-computed security level classification being more restricted than the user-generated security level classification, the one or more actions comprise causing a review of the dataset. ### Claim 3 (depends on 2) The apparatus of claim 2, wherein, following the review, the one or more actions comprise changing a current security level classification of the dataset. ### Claim 4 (depends on 1) The apparatus of claim 1, wherein, when the comparison results in the machine-computed security level classification being less restricted than the user-generated security level classification, the one or more actions comprise retraining the artificial intelligence model. ### Claim 5 (depends on 1) The apparatus of claim 1, wherein, when the machine-computed security level classification is one of one or more restricted security level classifications, the one or more actions comprise deleting the dataset from a device that downloaded the dataset. ### Claim 6 (depends on 1) The apparatus of claim 1, wherein, when the machine-computed security level classification is one of one or more restricted security level classifications, the one or more actions comprise encrypting the dataset. ### Claim 7 (depends on 1) The apparatus of claim 1, wherein, when the comparison results in the machine-computed security level classification being the same as the user-generated security level classification, the one or more actions comprise continuously monitoring a current security level classification of the dataset. ### Claim 8 (depends on 1) The apparatus of claim 1, wherein the one or more actions comprise tracking a location of the dataset. ### Claim 9 (depends on 1) The apparatus of claim 1, wherein the one or more actions comprise applying a severity level to the dataset based on the comparison, wherein the severity level is automatically selected from a plurality of severity levels and represents a degree of need for review of the dataset. ### Claim 10 (depends on 1) The apparatus of claim 1, wherein the given security-based data type comprises personally identifiable information. ### Claim 11 (depends on 1) The apparatus of claim 1, wherein the given security-based data type comprises entity-sensitive information. ### Claim 12 A method comprising: downloading a dataset from a data source; detecting whether or not the dataset contains at least a subset of data of a given security-based data type, wherein the detection is performed in accordance with an artificial intelligence model; generating a machine-computed security level classification for the dataset by automatically selecting one of a plurality of security level classifications based on the detection; comparing the machine-computed security level classification with a user-generated security level classification manually selected from the plurality of security level classifications; causing one or more actions to be taken based on the comparison; and creating a unique identifier for the dataset to track a location of the dataset, the unique identifier comprising a user identification and a timestamp; wherein, when the generated machine-computed security level classification is a restricted security level classification, the one or more actions comprise relocating the dataset based on the unique identifier from a device that downloaded the dataset to a secure storage location. ### Claim 13 (depends on 12) The method of claim 12, wherein, when the comparison results in the machine-computed security level classification being more restricted than the user-generated security level classification, the one or more actions comprise causing a review of the dataset. ### Claim 14 (depends on 12) The method of claim 12, wherein, when the comparison results in the machine-computed security level classification being less restricted than the user-generated security level classification, the one or more actions comprise retraining the artificial intelligence model. ### Claim 15 (depends on 12) The method of claim 12, wherein, when the machine-computed security level classification is one of one or more restricted security level classifications, the one or more actions comprise at least one of: deleting the dataset from a device that downloaded the dataset; and encrypting the dataset. ### Claim 16 (depends on 12) The method of claim 12, wherein, when the comparison results in the machine-computed security level classification being the same as the user-generated security level classification, the one or more actions comprise continuously monitoring a current security level classification of the dataset. ### Claim 17 (depends on 12) The method of claim 12, wherein the one or more actions comprise tracking a location of the dataset. ### Claim 18 (depends on 12) The method of claim 12, wherein the one or more actions comprise applying a severity level to the dataset based on the comparison, wherein the severity level is automatically selected from a plurality of severity levels and represents a degree of need for review of the dataset. ### Claim 19 A computer program product comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device cause the at least one processing device to: download a dataset from a data source; detect whether or not the dataset contains at least a subset of data of a given security-based data type, wherein the detection is performed in accordance with an artificial intelligence model; generate a machine-computed security level classification for the dataset by automatically selecting one of a plurality of security level classifications based on the detection; compare the machine-computed security level classification with a user-generated security level classification manually selected from the plurality of security level classifications; cause one or more actions to be taken based on the comparison; and create a unique identifier for the dataset to track a location of the dataset, the unique identifier comprising a user identification and a timestamp; wherein, when the generated machine-computed security level classification is a restricted security level classification, the one or more actions comprise relocating the dataset based on the unique identifier from a device that downloaded the dataset to a secure storage location. ### Claim 20 (depends on 19) The computer program product of claim 19, wherein, when the comparison results in the machine-computed security level classification being more restricted than the user-generated security level classification, the one or more actions comprise causing a review of the dataset.