Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
Applicant's arguments filed 06/08/2026 have been fully considered but they are not persuasive.
Applicant has amended claims 1,12 and 17. No new claims were added or cancelled. Currently claims 1-19 are pending in this application.
Applicant’s argument with respect to claims 1,12 and 17 have been considered but are moot in light of new grounds of rejection necessitated by applicant’s amendment listed below.
Claim Rejections - 35 USC § 103
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.
Claim(s) 1,4,7-13 and 17-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Mathur (US20180025174) in view of Jung (KR20230097713A) and in further view of Price (US20230315989) and Alfardan (US20240388551).
Regarding Claims 1, 12 and 17, Mathur discloses A method comprising: at a processor Mathur (Paragraph [0029])
monitoring usage of an electronic device to detect an interaction initiated Mathur (Paragraph [0028] E.N. Different capacities are allocated to the user based on their role. One of ordinary skill in the art can determine the device monitors who is using the electronic device in order to disclose information based on the user’s role.)
in response to detecting the interaction, determining, based on a usage policy or data privacy policy, a limitation on the information permitted to be provided Mathur (Paragraph [0028] E.N. Permission levels are determined based on the user’s hierarchy)
Mathur does not, but in related art, Jung discloses with an LLM AI Jung (Paragraph [0028] An application stored in the storage module is able to perform predefined operation, judgements, processing, and/or control operation.)
associated with [[an]] the LLM AI service provider; Jung (Paragraph [0003] E.N. OpenAI’s GPT-3 model is disclosed)
to the LLM AI service provider; Jung (Paragraph [0003] E.N. OpenAI’s GPT-3 model is disclosed)
identifying information submitted in a user interface as input to the LLM AI application to prompt a response generated by the LLM AI service provider; and Jung (Paragraph [0028] E.N. The decision simulation device receives specific text as input and generates a model. One of ordinary skill in the art can determine the information submitted is identified in order to generate a response.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur to incorporate the teachings of Jung because Mathur does not explicitly disclose LLM AI models and service provider which is taught by Jung. Incorporating the teachings of Jung to Mathur allows for the use of LLM AI models and service providers to monitor usage and provide some form of access control to data.
Mathur and Jung do not, but in related art, Price discloses and, in response to the scan, generating a first subset of the information that is permitted under the determined limitation and a second subset, different than the first subset, comprising the sensitive data; enabling provision to the LLM AI service provider of the first subset as the input and withholding the second subset by filtering the input to remove the second subset prior to transmission to the LLM AI service provider, wherein the withholding occurs in accordance with the limitation; Price (Paragraph [0056] E.N. Specified types of information in the patient information can be withheld as an input to training and/or use of the machine learning model.)
and optionally generalizing the first subset by removing names, values or substituting general wording for specific wording to further obfuscate sensitive information before it is transmitted. Price (Paragraph [0069] E.N. Tokenization is used to replace sensitive data with placeholder value for downstream processing.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose withholding information to be provided to a model which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for some sensitive information to be held back when used as an input to an LLM AI application/model to protect sensitive information of users. Mathur, Jung and Price do not, but in related art, Alfardan discloses within a firewall or cloud access security broker device positioned in a network architecture between an enterprise user device and an external large language model (LLM) artificial intelligence (AI) service provider, (Figure 1 E.N. A LLM Firewall/Gateway that is between the client and the LLM service is disclosed.)
scanning the input using a data loss prevention (DLP) engine to identify sensitive data that matches one or more policy defined criteria, (Paragraph [0054] E.N. As compared to Data Loss Prevention (DLP) engines, the LLM firewall implements functions and capabilities at the LLM level and goes beyond the capabilities of a DLP engine. For example, the LLM firewall system maintains a memory (context) for every client-LLM session/conversation to identify sophisticated multi-stage conversation attacks, such as those that try to circumvent any LLM model the system protects. Moreover, the LLM firewall tracks and monitors for malicious activities that an LLM might perform to jailbreak using an active session/conversation.)
including personal identifiable information (PII), protected health information (PHI), credit card numbers, enterprise secrets, source code, passwords, passkeys, financial data, merger and acquisition data, or other information not approved for external transmission, (Paragraph [0039] E.N. Rules are communicated to Client/Session Instances for enforcement. For example, RegEx expression can be created to match Personally Identifiable Information (PII) so as to block any messages with PII from reaching an LLM service.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung and in further view of Price to incorporate the teachings of Alfardan because Mathur, Jung and Price do not explicitly disclose firewall, LLM AI service provider which is taught by Alfardan. Incorporating the teachings of Alfardan to Mathur, Jung and Price to Alfardan allows for the firewall of the LLM AI service provider to block sensitive information from leaving the host server for better security.
Regarding Claim 12, Mathur further discloses A system comprising: a non-transitory computer-readable storage medium; and one or more processors (Paragraph [0030]).
coupled to the non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium comprises program instructions that, when executed on the one or more processors, cause the system to perform operations comprising (Paragraph [0030]).
Regarding Claim 17, Mathur further discloses A non-transitory computer-readable storage medium, storing instructions executable via one or more processors, (Paragraph [0030]).
Regarding Claims 4 and 13, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1 and the system of claim 12. Mathur and Jung do not, but in related art Price discloses wherein enabling provision to the LLM Al service provider of the first subset comprises, prior to transmission, applying a filtering operation to the information based on the determined limitation to remove a second subset that satisfies a predefined restriction criterion. (Paragraph 0056] E.N. Patient name and/or patient social security number is withheld from the model or otherwise redacted or removed from the database altogether. Doing so protects patient information privacy)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose withholding information to be provided to a model which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for some sensitive information to be held back when used as an input to an LLM AI application/model to protect sensitive information of users.
Regarding Claims 7, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1. Mathur and Jung do not, but in related art Price discloses further comprising automatically generalizing the first subset of information to remove sensitive information. (Paragraph [0069] E.N. Tokenization is used to replace sensitive data with placeholder values for downstream processing.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose generalizing subset of information to remove sensitive information which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for some sensitive information to be held back or generalized when used as an input to an LLM AI application/model to protect sensitive information of users.
Regarding Claim 8, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1. Mathur further discloses comprising identifying information received (Paragraph [0017] E.N. A processing device of a datastore system is configured to identify one or more grants of permission corresponding to one or more objects.)
Mathur does not, but in related art, Jung discloses from the LLM AI service provider and associating the received information with the input. (Paragraph [0003] E.N. OpenAI’s GPT-3 model is disclosed. One of ordinary skill in the art is able to determine the OpenAI’s GPT-3 model is able to receive input from a user.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur to incorporate the teachings of Jung because Mathur does not explicitly disclose LLM AI models and service provider which is taught by Jung. Incorporating the teachings of Jung to Mathur allows for the use of LLM AI models and service providers to monitor usage and provide some form of access control to data.
Regarding Claim 9, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 8. Mathur further discloses comprising watermarking the information received to associate the information received with a user who provided the input. (Paragraph [0028] E.N. Different users generally will have different capabilities with regards to accessing and modifying application and database information, depending on the user’s respective security or permission levels. One of ordinary skill in the art is able to determine some form of differentiation is done between the users accessing/modifying information.)
Regarding Claim 10, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1. Mathur does not, but in related art, Jung discloses as having been provided to the LLM AI service provider. (Paragraph [0003] E.N. OpenAI’s GPT-3 model is disclosed)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur to incorporate the teachings of Jung because Mathur does not explicitly disclose LLM AI models and service provider which is taught by Jung. Incorporating the teachings of Jung to Mathur allows for the use of LLM AI models and service providers to monitor usage and provide some form of access control to data.
Mathur and Jung do not, but in related art, Price discloses comprising logging the first subset of the information permitted under the determined limitation (Paragraph 0056] E.N. Patient name and/or patient social security number is withheld from the model or otherwise redacted or removed from the database altogether. One of ordinary skill in the art can determine the data is logged in a database.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose logging the first subset of less than all of the information which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for logging or storing data in a database to be used as input.
Regarding Claim 11, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1. Mathur and Jung do not, but in related art, Price discloses comprising initiating a right to be forgotten request to the LLM AI service provider based on determining that the first subset of the information permitted under the determined limitation was provided as input and contains data that satisfies a deletion criterion. (Paragraph [0056] E.N. Withheld information is removed from the database altogether. One of ordinary skill in the art can determine that the data is able to be removed (forgotten) altogether)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose a right to be forgotten request which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for removing data in a database to be used as input.
Regarding Claim 18, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1. Mathur and Jung do not, but in related art, Price discloses prior to transmission of the input to the LLM AI service provider, selectively processing the information to separate a first subset permitted for transmission from a second subset identified as sensitive or prohibited information. Price (Paragraph [0056] E.N. specified types of information in the patient information can be withheld as an input to training and/or use of the machine learning model because the types of information are deemed as bearing limited or reduced relevance to prediction accuracy. For example, patient name and/or patient social security number can be withheld from the machine learning model or otherwise redacted or removed from the patient database altogether. Doing so can protect patient information privacy while improving processing efficiency of training or using the machine learning model.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose withholding information to be provided to a model which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for some sensitive information to be held back when used as an input to an LLM AI application/model to protect sensitive information of users.
Regarding Claim 19, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 18. Mathur and Jung do not, but in related art, Price discloses wherein the selective processing reduces an amount of information transmitted to the LLM AI service provider by withholding the second subset prior to transmission. (Paragraph [0056] E.N. Doing so can protect patient information privacy while improving processing efficiency of training or using the machine learning model.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose selectively processing information to reduce the amount of information transmitted to the LLM AI service provider which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for some sensitive information to be held back when used as an input to an LLM AI application/model to protect sensitive information of users as well as improve efficiency of training or using the ML model.
Claim(s) 2-3,5-6 and 14-16 are rejected under 35 U.S.C. 103 as being unpatentable over Mathur (US20180025174) in view of Jung (KR20230097713A) and in further view of Price (US20230315989), Alfardan (US20240388551) and Desai (US20230370495).
Regarding Claim 2, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1. Mathur, Jung and Price do not, but in related art, Desai teaches: wherein the monitoring is performed by firewall that monitors incoming and outgoing network traffic. (Paragraph [0030] E.N. The cloud-based firewall provides Deep Packet Inspection and access controls across various ports and protocols as well as being application and user aware.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung and in further view of Price and Alfardan to incorporate the teachings of Desai because Mathur, Jung, Price and Alfardan do not explicitly disclose monitoring incoming and outgoing traffic by a firewall which is taught by Desai. Incorporating the teachings of Desai to Mathur, Jung, Price and Alfardan allow for the use of a firewall to see the packets coming in and out of the device when using a LLM AI application or service.
Regarding Claim 3, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1. Mathur, Jung and Price do not, but in related art, Desai teaches: wherein the monitoring is performed by a component positioned in a network architecture between one or more enterprise user devices and external cloud-based applications. (Figure 1A and Paragraph [0029] E.N. The cloud-based system offers access control and includes a cloud-based firewall. The cloud based system is between the user devices and cloud services.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung and in further view of Price and Alfardan to incorporate the teachings of Desai because Mathur, Jung, Price and Alfardan do not explicitly disclose monitoring incoming and outgoing traffic which is taught by Desai. Incorporating the teachings of Desai to Mathur, Jung, Price and Alfardan allow for the use of a firewall to see the packets coming in and out of the device when using a LLM AI application or service.
Regarding Claim 5 and 14, Mathur in view of Jung and in further view of Price and Alfardan discloses the method of claim 1 and the system of claim 12. Mathur and Jung do not, but in related art, Price discloses wherein enabling provision to the LLM AI service provider of the first subset of the information permitted under the determined limitation comprises, prior to transmission, (Paragraph [0056] E.N. Specified types of information in the patient information can be withheld as an input to training and/or use of the machine learning model)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose withholding information prior to transmission which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for some sensitive information to be held back when used as an input to an LLM AI application/model to protect sensitive information of users.
Mathur, Jung and Price do not, but in related art, Desai discloses applying a filtering operation using a data loss prevention (DLP) engine to identify sensitive data and remove a second subset that satisfies a predefined restriction criterion. (Paragraph [0032] E.N. The DPL uses standard and/or custom dictionaries to continuously monitor the users, including compressed and/or SSL encrypted traffic.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung and in further view of Price to incorporate the teachings of Desai because Mathur, Jung and Price do not explicitly disclose a data loss prevention engine which is taught by Desai. Incorporating the teachings of Desai to Mathur, Jung and Price allow for the use of a DLP engine to identity if any sensitive data is used as an input for a LLM AI application or service.
Regarding Claim 6, Mathur in view of Jung and in further view of Price, Alfardan and Desai discloses the method of claim 5. Mathur and Jung do not, but in related art, Price discloses wherein the sensitive data comprises personal identifiable information (PII), protected health information (PHI), credit card numbers, enterprise secrets, source code, passwords, passkeys, financial data, M&A data, or data not approved for use outside of an enterprise. (Paragraph 0056] E.N. Patient name and/or patient social security number is withheld from the model or otherwise redacted or removed from the database altogether. Doing so protects patient information privacy. One of ordinary skill in the art can determine the patient’s social security number is considered a PII and in some cases a PHI, the model uses data that is not considered sensitive so other sensitive data may also be withheld.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose the different types of sensitive data which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for sensitive data to be withheld as an input to a LLM AI application or service.
Regarding Claim 15, Mathur in view of Jung and in further view of Price and Alfardan discloses the system of claim 14. Mathur and Jung do not, but in related art, Price discloses wherein the sensitive data comprises personal identifiable information (PII), protected health information (PHI), credit card numbers, enterprise secrets, source code, passwords, passkeys, financial data, M&A data, or data not approved for use outside of an enterprise. (Paragraph 0056] E.N. Patient name and/or patient social security number is withheld from the model or otherwise redacted or removed from the database altogether. Doing so protects patient information privacy. One of ordinary skill in the art can determine the patient’s social security number is considered a PII and in some cases a PHI, the model uses data that is not considered sensitive so other sensitive data may also be withheld.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose the different types of sensitive data which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for sensitive data to be withheld as an input to a LLM AI application or service.
Regarding Claim 16, Mathur in view of Jung and in further view of Price and Alfardan discloses the system of claim 14. Mathur and Jung do not, but in related art Price discloses further comprising automatically generalizing the first subset of information to remove sensitive information. (Paragraph [0069] E.N. Tokenization is used to replace sensitive data with placeholder values for downstream processing.)
Therefore, it would be obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention to have modified Mathur in view of Jung to incorporate the teachings of Price because Mathur and Jung do not explicitly disclose generalizing subset of information to remove sensitive information which is taught by Price. Incorporating the teachings of Price to Mathur and Jung allows for some sensitive information to be held back or generalized when used as an input to an LLM AI application/model to protect sensitive information of users.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to AAYUSH ARYAL whose telephone number is (571)272-2838. The examiner can normally be reached 8:00 a.m. - 5:30 p.m..
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, Amir Mehrmanesh can be reached at (571) 270-3351. 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.
/AAYUSH ARYAL/Examiner, Art Unit 2435
/AMIR MEHRMANESH/Supervisory Patent Examiner, Art Unit 2435
Prosecution Insights