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 .
Status of the Claims
Claims 1-2, 7, and 12 are amended. Claims 9-11 have been cancelled. Claims 1-8 and 12-13 are currently pending and have been considered by the Examiner.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-8 and 12-13 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
In claim 1, the limitation in the final 2 lines renders the claim indefinite. It is unclear what constitutes the output data of “the at least one AI model” being processed by the generic AI model, since the claim discloses above in lines 9-10 that the at least one AI model includes the generic AI model. The limitation “the at least one AI model of the plurality of right-specialized AI models” is unclear since the claim discloses in lines 9-10 that the AI model itself includes the right-specialized AI models. Therefore, it is unclear if said AI model comprises multiple right-specialized AI models or if said AI model is one of the right-specialized AI models. Examiner treats claim 1 to mean the data is output through the generic AI model.
Claims 2-8 and 13 are rejected for failing to cure the deficiencies of claim 1.
Claim 12 recites the same indefinite limitations as claim 1 and is therefore rejected for at least the same reasons.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-13 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Claims 1-11 recite a method, claim 12 recites an access control device comprising a processor (a system), and claim 13 recites a non-transitory computer-readable recording medium (a product). A method, a system, and a product each falls within one of the four statutory categories of patent eligible subject matter.
Claim 1
Step 2A Prong 1: Filtering information available for the at least one
Step 2A Prong 2: At least one artificial intelligence (AI) model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
Acquiring output data of the at least one AI model based on the filtered data, wherein the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The limitations “learning information” and “requested information” indicate the filtered data is used for learning or inference.
The at least one Al model includes a generic Al model and a plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h).
Specific filtered data is based on at least one of a specific access right or a specific access control policy among the plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h).
The specific filtered data is input to a specific right-specialized Al model which is trained based on at least one of the specific access right or the specific access control policy amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, alone or in combination, do not integrate the abstract ideas into a practical application as they are mere insignificant extra solution activities as disclosed in combination with generic computer functions and a mere field of use that are implemented to perform the abstract ideas disclosed above. The claim is directed to an abstract idea.
Step 2B: At least one artificial intelligence (AI) model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
Acquiring output data of the at least one AI model based on the filtered data, wherein the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The limitations “learning information” and “requested information” indicate the filtered data is used for learning or inference.
The at least one Al model includes a generic Al model and a plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h).
Specific filtered data is based on at least one of a specific access right or a specific access control policy among the plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h).
The specific filtered data is input to a specific right-specialized Al model which is trained based on at least one of the specific access right or the specific access control policy amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, in combination with the abstract ideas, are not sufficient to amount to significantly more than the abstract ideas as they are well-understood, routine and conventional activities as disclosed in combination with generic computer functions and a mere field of use that are implemented to perform the abstract ideas disclosed above. The claim is not patent eligible.
Claim 2 incorporates the rejection of claim 1.
Step 2A Prong 1: The abstract ideas of claim 1 are incorporated.
Step 2A Prong 2 and Step 2B: The learning information corresponds to learning allowable information associated with access right meta information corresponding to at least one of the access right or the access control policy among the available information amounts to a mere field of use and technological environment under MPEP 2106.05(h).
A learning process for the at least one AI model is performed based on the learning information amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The claim is not patent eligible.
Claim 3 incorporates the rejection of claim 2.
Step 2A Prong 1: The abstract ideas of claim 2 are incorporated.
Step 2A Prong 2 and Step 2B: The learning allowable information is applied to fine-tuning learning for the at least one AI model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The claim is not patent eligible.
Claim 4 incorporates the rejection of claim 2.
Step 2A Prong 1: The abstract ideas of claim 2 are incorporated.
Step 2A Prong 2 and Step 2B: The access right meta information includes information on at least one of a creator, a modifier, an owner, a manager, a management group, a right identifier, a right level or a security level amounts to a mere field of use and technological environment under MPEP 2106.05(h). The claim is not patent eligible.
Claim 5 incorporates the rejection of claim 1.
Step 2A Prong 1: The abstract ideas of claim 1 are incorporated.
Step 2A Prong 2 and Step 2B: The requested information is input to the at least one AI model learned based on the learning information included in the filtered data or at least one AI model learned based on general learning information not included in the filtered data amounts to invoking computers merely as a tool to perform an existing process under MPEP 2106.05(f).
The output data is acquired through an inference process based on the requested information amounts to invoking computers merely as a tool to perform an existing process under MPEP 2106.05(f). The claim is not patent eligible.
Claim 6 incorporates the rejection of claim 1.
Step 2A Prong 1: The abstract ideas of claim 1 are incorporated.
Step 2A Prong 2 and Step 2B: The requested information includes at least one of user information, direct input information or additional input information amounts to a mere field of use and technological environment under MPEP 2106.05(h). The claim is not patent eligible.
Claim 7 incorporates the rejection of claim 6.
Step 2A Prong 1: The abstract ideas of claim 6 are incorporated.
Step 2A Prong 2 and Step 2B: The additional input information is based on extension allowable information associated with access right meta information corresponding to the access control policy among the available information amounts to a mere field of use and technological environment under MPEP 2106.05(h).
The access control policy is based on at least one of the user information or the direct input information amounts to a mere field of use and technological environment under MPEP 2106.05(h). The claim is not patent eligible.
Claim 8 incorporates the rejection of claim 7.
Step 2A Prong 1: The abstract ideas of claim 7 are incorporated.
Step 2A Prong 2 and Step 2B: The additional input information includes at least one of knowledge information related to the direct input information or a keyword generated based on the direct input information amounts to a mere field of use and technological environment under MPEP 2106.05(h). The claim is not patent eligible.
Claim 12
Step 2A Prong 1: Filter information available for the at least one
Step 2A Prong 2: An access control device for at least one artificial intelligence (AI) model, the device comprising: a transceiver, a memory, a user interface, and a processor amount to generic computer components for applying the abstract ideas on a generic computer under MPEP 2106.05(f). An AI model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
Store filtered data in the memory amounts to insignificant post-solution activities under MPEP 2106.05(g).
Acquire output data of the at least one AI model based on the filtered data amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
Store output data in the memory amounts to insignificant post-solution activities under MPEP 2106.05(g).
The limitation “the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model” indicates the filtered data is used for learning or inference, which amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The at least one Al model includes a generic Al model and a plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h).
Specific filtered data is based on at least one of a specific access right or a specific access control policy among the plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h).
The specific filtered data is provided to a specific right-specialized Al model which is trained based on at least one of the specific access right or the specific access control policy amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, alone or in combination, do not integrate the abstract ideas into a practical application as they are mere insignificant extra solution activities as disclosed in combination with generic computer functions and a mere field of use that are implemented to perform the abstract ideas disclosed above. The claim is directed to an abstract idea.
Step 2B: An access control device for at least one artificial intelligence (AI) model, the device comprising: a transceiver, a memory, a user interface, and a processor amount to generic computer components for applying the abstract ideas on a generic computer under MPEP 2106.05(f). An AI model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
Store filtered data in the memory is a well-understood, routine, conventional activity recognized by the courts under MPEP 2106.05(d)(II).
Acquire output data of the at least one AI model based on the filtered data amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
Store output data in the memory is a well-understood, routine, conventional activity recognized by the courts under MPEP 2106.05(d)(II).
The limitation “the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model” indicates the filtered data is used for learning or inference, which amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, in combination with the abstract ideas, are not sufficient to amount to significantly more than the abstract ideas as they are well-understood, routine and conventional activities as disclosed in combination with generic computer functions and a mere field of use that are implemented to perform the abstract ideas disclosed above. The claim is not patent eligible.
Claim 13 recites a product which implements the same features as the method of claim 1 and is therefore rejected for at least the same reasons.
In Step 2A Prong 2 and Step 2B, a non-transitory computer-readable recording medium storing a computer program for executing a method at a computer amount to generic computer components for applying the abstract ideas on a generic computer under MPEP 2106.05(f). The claim is not patent eligible.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-8 and 12-13 are rejected under 35 U.S.C. 103 as being unpatentable over Ramakrishnan et al. (US 20240177049 A1, cited in PTO-892 issued 06/25/2026) in view of O’Connor et al. (US 20220092404 A1).
Regarding claim 1, Ramakrishnan teaches: An access control method for at least one artificial intelligence (AI) model, the method comprising: filtering information available for the at least one AI model based on at least one of an access right or an access control policy and acquiring filtered data for the at least one AI model; and ([0016], [0019], lines 1-8 and [0042], lines 1-10 teaches fine-tuning a pre-trained model using a confidential tuning data set. The data set may have user data access restrictions, which may allow a particular user to tune the pre-trained model based on the user’s own confidential data while preventing other users from accessing the confidential data.)
acquiring output data of the at least one AI model based on the filtered data, ([0014], lines 1-17, [0016], and [0039]-[0041], where output data corresponds to an inference result outputted by the tuned model, and filtered data corresponds to confidential tuning data. Since the inference result is based on the model’s learned parameters, and the parameters had been learned based on the confidential tuning data, the inference result is therefore based on the confidential tuning data.)
wherein the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model, ([0019], lines 1-8 and [0042], lines 1-10, where a confidential tuning data set corresponds to the limitation “learning information for the at least one AI model” as claimed.)
wherein specific filtered data is based on at least one of a specific access right or a specific access control policy among the plurality of right-specialized Al models, ([0019], lines 1-8 and [0042], lines 1-10)
wherein the specific filtered data is input to a specific right-specialized Al model which is trained based on at least one of the specific access right or the specific access control policy, ([0016], [0019], lines 1-8, and [0042], lines 1-10 discloses inputting confidential tuning data to a pre-trained machine learning model in order to fine-tune it.)
wherein the output data
However, Ramakrishnan does not explicitly teach: wherein the at least one Al model includes a generic Al model and a plurality of right-specialized Al models,
the output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model.
But O’Connor teaches: wherein the at least one Al model includes a generic Al model and a plurality of
the output data of the at least one Al model of the plurality of
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated O’Connor’s architecture comprising a data classifier neural network and a plurality of student neural networks into Ramakrishnan, where each student neural network is one of the plurality of tuned models 347. A motivation for the combination is the architecture would streamline data processing by automatically clustering input data by access right and then routing it to a correct fine-tuned neural network.
Regarding claim 2, the combination of Ramakrishnan and O’Connor teaches: The method of claim 1, wherein:
Ramakrishnan teaches: the learning information corresponds to learning allowable information associated with access right meta information corresponding to at least one of the access right or the access control policy among the available information, and ([0019], lines 1-8 and [0042], lines 1-10 teaches allowing permitted users to access the tuning data set while preventing other users from accessing it. The limitation “access right meta information” includes location and access credentials to obtain the data set.)
a learning process for the at least one AI model is performed based on the learning information. ([0016] where a learning process includes fine-tuning.)
Regarding claim 3, the combination of Ramakrishnan and O’Connor teaches: The method of claim 2, wherein:
Ramakrishnan teaches: the learning allowable information is applied to fine-tuning learning for the at least one AI model. ([0016])
Regarding claim 4, the combination of Ramakrishnan and O’Connor teaches: The method of claim 2, wherein:
Ramakrishnan teaches: the access right meta information includes information on at least one of a creator, a modifier, an owner, a manager, a management group, a right identifier, a right level or a security level. ([0019], lines 1-8 teaches the limitation “access right meta information includes information on… a right level”.)
Regarding claim 5, the combination of Ramakrishnan and O’Connor teaches: The method of claim 1, wherein:
Ramakrishnan teaches: the requested information is input to the at least one AI model learned based on the learning information included in the filtered data or at least one AI model learned based on general learning information not included in the filtered data, ([0014], lines 7-17 and [0016], lines 1-14 teaches performing inference on a confidentially tuned model. The feature of inputting requested information to the AI model corresponds to inputting new, unseen data to the confidentially tuned model.)
the output data is acquired through an inference process based on the requested information. ([0014], lines 13-17)
Regarding claim 6, the combination of Ramakrishnan and O’Connor teaches: The method of claim 1, wherein:
Ramakrishnan teaches: the requested information includes at least one of user information, direct input information or additional input information. ([0014], lines 7-17 and [0016], lines 1-14 teaches performing inference on a confidentially tuned model. The limitation “direct input information” corresponds to inputting new, unseen data to the confidentially tuned model.)
Regarding claim 7, the combination of Ramakrishnan and O’Connor teaches: The method of claim 6, wherein:
Ramakrishnan teaches: the additional input information is based on extension allowable information associated with access right meta information corresponding to the access control policy among the available information, and the access control policy is based on at least one of the user information or the direct input information. ([0039]-[0041] teaches a model user account may be associated with the confidentially tuned model, and only that model user account may send inference requests to the confidentially tuned model. The limitation “additional input information” includes an access control list that associates the model user account with the confidentially tuned model.)
Regarding claim 8, the combination of Ramakrishnan and O’Connor teaches: The method of claim 7, wherein:
Ramakrishnan teaches: the additional input information includes at least one of knowledge information related to the direct input information or a keyword generated based on the direct input information. ([0014], lines 13-17 and [0039]-[0041] teaches a model user account may be associated with the confidentially tuned model, and only that user account may send inference requests to the confidentially tuned model. The limitation “knowledge information related to the direct input information” includes knowledge that a model user account is associated with the confidentially tuned model, and only that account may send inference requests to the model.)
Regarding claim 12, Ramakrishnan teaches: An access control device for at least one artificial intelligence (AI) model, the device comprising: ([0059], lines 1-5)
a transceiver; ([0065] where a transceiver is a network interface)
a memory; ([0063], lines 1-2)
a user interface; and ([0060] on page 8, col. 1, lines 3-5. A user interface is a display including a computer monitor.)
a processor, wherein the processor is configured to: ([0061], lines 1-6)
filter information available for the at least one AI model based on at least one of an access right or an access control policy, acquire filtered data for the at least one AI model and store it in the memory; and ([0016], [0019], lines 1-8 and [0042], lines 1-10 teaches fine-tuning a pre-trained model using a confidential tuning data set. The data set may have user data access restrictions, which may allow a particular user to tune the pre-trained model based on the user’s own confidential data while preventing other users from accessing the confidential data. [0067], lines 1-5 and 7-9 teaches storing data in memory, which would include confidential tuning data set.)
acquire output data of the at least one AI model based on the filtered data and store it in the memory, ([0014], lines 1-17, [0016], [0039]-[0041], and [0067], lines 1-5 and 7-9, where output data corresponds to an inference result outputted by the tuned model, and filtered data corresponds to confidential tuning data. Since the inference result is based on the model’s learned parameters, and the parameters had been learned based on the confidential tuning data, the inference result is therefore based on the confidential tuning data.)
wherein the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model, ([0019], lines 1-8 and [0042], lines 1-10, where a confidential tuning data set corresponds to the limitation “learning information for the at least one AI model” as claimed.)
wherein specific filtered data is based on at least one of a specific access right or a specific access control policy among the plurality of right-specialized AI models, ([0019], lines 1-8 and [0042], lines 1-10)
wherein the specific filtered data is provided to a specific right-specialized AI model which is trained based on at least one of the specific access right or the specific access control policy, ([0016], [0019], lines 1-8, and [0042], lines 1-10 discloses inputting confidential tuning data to a pre-trained machine learning model in order to fine-tune it.)
wherein the output data
However, Ramakrishnan does not explicitly teach: wherein the at least one Al model includes a generic Al model and a plurality of right-specialized Al models,
the output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model.
But O’Connor teaches: wherein the at least one Al model includes a generic Al model and a plurality of is data classifier neural network 107 and “a plurality of… specialized AI models” are student neural networks 106.)
the output data of the at least one Al model of the plurality of
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated O’Connor’s architecture comprising a data classifier neural network and a plurality of student neural networks into Ramakrishnan, where each student neural network is one of the plurality of tuned models 347. A motivation for the combination is the architecture would streamline data processing by automatically clustering input data by access right and then routing it to a correct fine-tuned neural network.
Claim 13 is recites a product which implements the same features as the method of claim 1 as is therefore rejected for at least the same reasons.
Ramakrishnan teaches: A non-transitory computer-readable recording medium storing a computer program for executing a method at a computer ([0058], lines 3-8)
Response to Arguments
Below are the Examiner’s response to the Applicant’s arguments filed on 09/25/2026.
Applicant’s Arguments Under 35 U.S.C. 101 – Step 2A: On page 8 of the remarks, regarding Step 2A Prong 1, the Applicant argues the previous Office Action did not analyze claim 1 based on the actual claim language, and the Applicant argues the claims merely involve a judicial exception. On page 9 of the remarks, regarding Step 2A Prong 2, the Applicant argues that filtering information, acquiring filtered data, and acquiring output data improves the overall capability of the AI model.
Examiner’s Response: Applicant's arguments have been fully considered but they are not persuasive. With respect to Step 2A Prong 1, the analysis of claim 1 is based on the actual claim language “filtering information available for the at least one… model based on at least one of an access right or an access control policy and acquiring filtered data for the at least one… model”. This is a method of organizing human activity. The sub-grouping “managing personal behavior or relationships or interactions between people” includes following rules or instructions. A person with a “medium” level of access rights would be allowed to access documents having “medium” or “low” right levels. A person with a “low” level of access rights would be allowed to access documents having “low” right levels and blocked from accessing documents having “medium” right levels. Therefore, filtering by access rights is used to manage personal behavior. Specification paragraphs [0076]-[0077] were cited because they explain how access rights are used to manage personal behavior. For these reasons, Examiner respectfully disagrees with the argument that claim 1 merely involves a judicial exception.
With respect to Step 2A Prong 2, the Examiner respectfully disagrees that the limitations of claim 1 improve the overall capability of the AI model. First, the limitation of filtering information available for the at least one model based on at least one of an access right or an access control policy and acquiring filtered data for the at least one model is a method of organizing human activity (an abstract idea). The judicial exception alone cannot provide the improvement, and an improvement in the abstract idea itself is not an improvement in technology, according to MPEP 2106.05(a) and (a)(II).
The claim limitations do not explain how the operation, structure, or capabilities of the AI model are improved. The remarks argue that the claimed features improve “the overall capability of the AI model” but neither the remarks nor specification paragraphs [0024], [0066]-[0067] clearly explain this capability.
The feature of at least one artificial intelligence (AI) model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). Acquiring output data of the at least one AI model based on the filtered data, wherein the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The limitations “learning information” and “requested information” indicate the filtered data is used for learning or inference. The at least one Al model includes a generic Al model and a plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h). Specific filtered data is based on at least one of a specific access right or a specific access control policy among the plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h). The specific filtered data is input to a specific right-specialized Al model which is trained based on at least one of the specific access right or the specific access control policy amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, alone or in combination, do not integrate the abstract ideas into a practical application as they are mere insignificant extra solution activities as disclosed in combination with generic computer functions and a mere field of use that are implemented to perform the abstract ideas disclosed above. The claim is directed to an abstract idea.
Applicant’s Arguments Under 35 U.S.C. 101 – Step 2B: On page 9 of the remarks, the Applicant argues the claims recite a technical implementation for solving a specific technical problem, that the claims recite a specific combination of operations where such combination is not taught by the cited references, and that the limitations of claim 1 achieves technical improvements in AI model functionality significantly more than the abstract ideas.
Examiner’s Response: Applicant's arguments have been fully considered but they are not persuasive. It is noted that comparing the claim limitations to the cited references is used for claim analysis under 35 U.S.C. 102 and 103, not under 101.
Examiner respectfully disagrees with the Applicant’s arguments. At least one artificial intelligence (AI) model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). Acquiring output data of the at least one AI model based on the filtered data, wherein the filtered data includes at least one of learning information for the at least one AI model or requested information for the at least one AI model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The limitations “learning information” and “requested information” indicate the filtered data is used for learning or inference. The at least one Al model includes a generic Al model and a plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h). Specific filtered data is based on at least one of a specific access right or a specific access control policy among the plurality of right-specialized Al models amounts to a mere field of use and technological environment under MPEP 2106.05(h). The specific filtered data is input to a specific right-specialized Al model which is trained based on at least one of the specific access right or the specific access control policy amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f). The output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model amounts to mere instructions to apply the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, in combination with the abstract ideas, are not sufficient to amount to significantly more than the abstract ideas as they are well-understood, routine and conventional activities as disclosed in combination with generic computer functions and a mere field of use that are implemented to perform the abstract ideas disclosed above. The claim is not patent eligible.
Applicant’s Arguments Under 35 U.S.C. 102: On page 10, the Applicant argues that Ramakrishnan does not disclose the specific access right and/or specific access control policy of claim 1, which are related to the specific filtered data that is filtered from information available for the plurality of right-specialized Al models. The Applicant argues that Ramakrishnan does not disclose that common available information is differently filtered and used by different right-specialized Al models according to claim 1. The Applicant argues that Ramakrishnan does not disclose “the output data of the at least one Al model of the plurality of right-specialized Al models are integrated or processed by the generic Al model” as required by amended claim 1.
Examiner’s Response: Applicant's arguments have been fully considered but they are not persuasive. Ramakrishnan discloses “specific access right” as a user data access restriction. The system stores confidential data belonging to each user. When a user desires to tune a pre-trained model on the user’s own confidential data, the system performs this tuning while blocking access to confidential data that belongs to the other users. Therefore, the system filters out all confidential data except for data belonging to the user. See paragraphs [0016], [0019], lines 1-8, [0038]-[0041], and [0042], lines 1-10.
Furthermore, the claim does not positively recite that common information is filtered, input to a trained right-specialized AI model, and used to make a prediction. The claim limitations in lines 6-8 are broad. The filtered data can include learning information (training data), and the output data is merely “based on the filtered data” in any way. This Office Action treats the limitations in lines 6-8 to mean using confidential data to tune a pre-trained model by adjusting model parameters, and outputting an inference result on new data based on the model parameters, which were adjusted based on the confidential data. See paragraphs [0016], [0019], lines 1-8, [0038]-[0041], and [0042], lines 1-10.
The Examiner has considered the Applicant’s arguments with respect to the limitations of claim 1, lines 9-10 and 15-16. These limitations render claim 1 indefinite for the reasons indicated in the rejection of claim 1 under 35 U.S.C. 112. Furthermore, the arguments regarding one AI model containing multiple AI models 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.
Conclusion
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 Asher H. Jablon whose telephone number is (571)270-7648. The examiner can normally be reached Monday - Friday, 9:00 am - 6:00 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, Abdullah Al Kawsar can be reached at (571)270-3169. 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.
/A.H.J./Examiner, Art Unit 2127
/ABDULLAH AL KAWSAR/Supervisory Patent Examiner, Art Unit 2127
