DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This office action is in response to communication filed on April 27, 2026.
Status of claims within the present application:
Claims 1 – 2 are pending.
Claims 1 – 2 are amended.
Response to Arguments
Regarding claim 1 that was objected to, Applicant’s remark, see page [4], filed on April 27, 2026, has been considered and are persuasive. Therefore, the objection is withdrawn.
Regarding claim 1 that was interpreted under 35 U.S.C. 112(f), see page [4 – 5], filed on April 27, 2026, has been considered and are persuasive. Therefore, the interpretation is withdrawn.
Regarding claims 1 – 2 that were rejected under 35 U.S.C. 103 as being unpatentable over US 20230385451 A1 to Lockhart, III et al., (hereinafter, “Lockhart”) in view of US 20230006845 A1 to Leedom, JR., (hereinafter, “Leedom”), Applicant’s remark, see page [4], filed on April 27, 2026, has been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1 – 2 are rejected under 35 U.S.C. 103 as being unpatentable over US 20230385451 A1 to Lockhart, III et al., (hereinafter, “Lockhart”) in view of US 20230006845 A1 to Leedom, JR., (hereinafter, “Leedom”) and US 12197483 B1 to Shmukler et al., (hereinafter, “Shmukler”).
Regarding claim 1, Lockhart teaches an AI learning system comprising: at least one memory storing instructions; and at least one processor that is configured to execute the instructions to: sort, from among AI learning data, personal information when creating an AI model, according to prescribed criteria; [Lockhart, para. 11 discloses the system may crawl one or more identified web sites to extract data. The system may process the extracted data, searching the extracted data to identify patterns representing portions of PII data. The scraped data may be provided to an Artificial Intelligence (AI) engine for processing against particular rules to verify PII data or may be elevated to an administrator for review. Para. 14 discloses the compromised PII exchange system and method may be configured to not only search (and crawl) an initially identified set of websites based on an list of URLs (e.g., URLs for websites on the dark web) that have been provided to the system, but also automatically search additional websites identified by URLs detected in the initially identified websites.] encrypt the sorted personal information; [Lockhart, para. 39 discloses the compromised identity exchange system may disassociate and encrypt the PII data from an at-risk entity if the at risk entity did not perform the disassociation and may communicate the encrypted data to one or more of the compromised entities in response to the query. The compromised identity exchange system may receive results from the one or more entities in response to the queries where a match was made to a full PII identity or disassociated identity elements.] wherein secret calculations are performed on the encrypted personal information included in the stored Al learning; [Lockhart, para. 46 discloses a consumer or an at-risk entity may want to determine if its data may correspond in some way to the data that was exposed. The consumer or at-risk entity may communicate at least a portion of its PII data to the compromised PII exchange system 102 for comparison against the compromised PII data 122. In some embodiments, the portion of the PII data may be disassociated and encrypted prior to transmission. The compromised PII exchange system 102 may re-encrypt the PII data in the same manner as the PII data stored in the compromised PII data 122 and may compare the re-encrypted PII data from the source to the compromised PII data 122. The compromised PII exchange system 102 may return data related to the results of the comparison. Para. 47 discloses the data returned may include a risk assessment score based on the results of the comparison. For example, if the data corresponds to PII data that has previously been identified in a fraudulent transaction, or that the compromised entity data breach is actively being used in fraudulent ways, the risk assessment score may be high. In another example, if the data results correspond to a low-risk event (such as a lost laptop computer) or an older event with no known harm, the risk assessment score may be lower], but Lockhart does not teach hold the AI learning data including the encrypted personal information in a blockchain; computationally process the stored AI learning data to create the AI model; output the AI model; and set the prescribed criteria by adjusting a type and combination of one or more hyperparameters related to a processing amount of secure computations and a secrecy of the personal information during the Al learning process, so as to reduce amount of calculation processing.
However, Leedom does teach hold the AI learning data including the encrypted personal information in a blockchain. [Leedom, para. 77 discloses The operating systems within each device are designed to aid in authenticating the user including time variable encryption keys generated by each of the paired hardware components to allow secure encryption of the personal data arranged to be stored immutably in the associated personal blockchain or as blocks in a blockchain, serving other users and/or purposes, wherein only the user can enable the personal computer device(s) to grant access to those blocks containing the immutably stored personal data.]
Therefore, it would have been obvious to one of ordinary skill within the art before the effective filling date to combine Leedom’s system with Lockhart’s system, with a motivation for user transactions, such as transactions successively enabled over time by the user's smartphone and/or paired smartwatch will generate user data that may be encrypted within blocks of the user's blockchain. The immutably stored personal data will be date stamped and stored in successive blocks of the associated personal blockchain. By creating and storing parts of each encryption key in the respective paired personal computers, the encryption of personal data for storage in the personal blockchain may be undertaken in a way that requires the respective paired personal computers to be in sufficient proximity for the DApp (smartcontract and blockchain) to operate properly. [Leedom, para. 75]
However, Lockhart in view of Leedom does not teach computationally process the stored AI learning data to create the AI model; output the AI model; and set the prescribed criteria by adjusting a type and combination of one or more hyperparameters related to a processing amount of secure computations and a secrecy of the personal information during the Al learning process, so as to reduce amount of calculation processing, but Shmukler does teach computationally process the stored AI learning data to create the AI model, [Shmukler, col. 18 lines 18 – 33 discloses a computerized method comprising: (a) performing a deterministic rule-based search, in a plurality of stored documents, for Personally Identifiable Information (PII) data-items; (b) if the deterministic rule-based search indicates that a particular document contains a PII data-item then: (b1) extracting a textual snippet from said particular document, wherein the textual snippet surrounds said PII data-item; (b2) adding an entirety of said particular document to a Machine Language classification dataset, that is utilized for training a Machine Learning (ML) engine to classify documents as either containing PII or non-containing PII; (b3) adding said textual snippet, and not the entirety of said particular document, to a second dataset utilized for training a Large Language Model (LLM) to find PII data-items in documents for performing Named Entity Recognition (NER) in said documents.] output the AI model; [Shmukler, col. 8 lines 35 – 48 discloses to utilize the pre-trained LLM, for example, a linear layer is incorporated with SoftMax activation that is connected to the final hidden layer (or via other suitable means; or other normalized exponential function, or other suitable activation function that scales numbers or logits into probabilities and outputs a vector with the probabilities of each possible outcome; or other function that generalizes logit-based or logistic function to multiple dimensions; or a suitable multinomial logistic regression function or sets-of-functions; or other suitable activation function and particularly a last activation function of a neural network (NN) that normalizes the output of the NN to a probability distribution over predicted output classes), and the LLM is fit (block 224).] and set the prescribed criteria by adjusting a type and combination of one or more hyperparameters related to a processing amount of secure computations and a secrecy of the personal information during the Al learning process, so as to reduce amount of calculation processing. [Shmukler, col. 8 lines 48 – 62 discloses the step of fitting the LLM may include one or more iterations in which parameters or keys of the LLM are modified or adjusted, to increase its accuracy, to prevent or cure over-fitting, to prevent or cure under-fitting, to reduce error level, and/or to otherwise improve the matching between outputs of the LLM and “ground-truth” results that are known to be correct, and/or to otherwise modify or adjust LLM parameters to reduce the difference between the data that it outputs and the ground-truth data (e.g., using Weber's Law and by finding the best fitting Weber fraction). In some embodiments, for example, a pre-tokenization (or pre-vectorization) process may optionally be performed on some documents or document-portions or document snippets, to prevent or reduce over-fitting;]
Therefore, it would have been obvious to one of ordinary skill within the art before the effective filling date to combine Shmukler’s system with modified Lockhart’s system, with a motivation for upon finding a PII data-item in a document and then extracting the relevant Document Snippet, this particular document (or a copy thereof) and this particular Document Snippet are collected by (or, are routed to, or transferred to) a Local Collector unit, for further training/re-training/updating of a LLM that is utilized by the LLM engine. [Shmukler, col. 8 lines 9 – 15]
As per claim 2, modified Lockhart teaches the AI learning system according to claim 1, wherein the at least one processor is configured to execute the instructions to: set the prescribed criteria and sort the personal information by AI learning using the degree of increase/decrease in the amount of calculation processing and the degree to which the personal information should be secret as AI learning data. [Lockhart, para. 39 discloses the compromised identity exchange system may receive results from the one or more entities in response to the queries where a match was made to a full PII identity or disassociated identity elements. Each match returned can include information about the data breach, which may consists of the date of the breach, the size/volume of the breach, a code indicating how the data was lost or stolen, among other attributes. In addition to these attributes, attributes associated with the consumer may also be used to measure risk. These attributes might include the number and severity of data breaches a consumer has been involved with, the location of the consumer, the event, if any, that is triggering the risk assessment, among other things. Additionally, participating at-risk entities' reported fraud data will be used to identify fraud rates within every compromised entity's compromised file, as well as attributes will be generated that reflect location of fraud, fraud linkages to email, physical address, phone number or other identity elements.]
Conclusion
Pertinent prior art made of record however not relied upon:
US 20240089081 A1 to Kushnir et al.
“An example system includes a processor to compute a tensor of indicators indicating a presence of partial sums in an encrypted vector of indicators. The processor can also securely reorder an encrypted array based on the computed tensor of indicators to generate a reordered encrypted array.”
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Phuc Pham whose telephone number is (571)272-8893. The examiner can normally be reached Monday - Thursday 7:30 AM - 4:30 PM; Friday 8:00 AM - 12:00 PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Linglan Edwards can be reached at (571) 270-5440. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/P.P./Patent Examiner, Art Unit 2408
/LINGLAN EDWARDS/Supervisory Patent Examiner, Art Unit 2408
Prosecution Insights