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 Application
This final office action is in response to the arguments filed 2/18/2026. Claims 1, 6-8, 13-15, and 19-20 have been amended. Claims 1-20 are currently pending and have been examined below.
Claim Rejections – 35 U.S.C. 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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Per step 1 of the eligibility analysis set forth in MPEP § 2106, subsection III, the claims are directed towards a process, machine, or manufacture.
Per step 2A Prong One, Claim 1 recites specific limitations which fall within at least one of the groupings of abstract ideas enumerated in MPEP 2106.04(a)(2) as follows:
determining, based on groups of online feedback information on previously displayed advertisements in respective user segments, performance metrics associated with the previously displayed advertisements with respect to the respective user segments;
generating segment-sensitive training data based on the user segments and their corresponding performance metrics, wherein segment-sensitive training data created for each of the user segments comprises information about the user segment, assets and features of the previously displayed advertisements to users in the user segment, and the performance metrics associated with the user segment;
generates segment-sensitive advertisement assets with respect to users in the corresponding user segment based on segment-dependent characteristics;
receiving, base advertisement information associated with an advertisement for a product, wherein the base advertisement information specifies a base image to characterize features of the product;
modifying, for the user segments a subgroup of the features exhibited in the base image to create respective segment-sensitive advertisement assets of the product;
the different image assets of the product used to form different asset combinations each of which can be used to display the advertisement for the product.
As noted above, these limitations fall within at least one of the groupings of abstract ideas enumerated in the MPEP 2106.04(a)(2). Specifically, these limitations fall within the group Certain Methods of Organizing Human Activity (i.e., fundamental economic principles or practices (including hedging, insurance, mitigating risk); commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations); managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions). That is – the limitations recited above describe modifying parts of a base ad asset to create alternate versions for different user segments based on the performance of previously displayed advertisements with respective user segments which is an advertising / marketing activity that falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas. Accordingly claim 1 recites an abstract idea.
Per step 2A Prong 2, the Examiner finds that the judicial exception is not integrated into a practical application. Claim 1 recites the additional limitations of:
[receiving] from a data storage, [base advertisement information] and storing, into the data storage [the different image assets];
obtaining, via machine learning, based on the segment-sensitive training data, generative artificial intelligence (Al) models with respect to the user segments, wherein each of the generative AI models with respect to the user segments wherein each of the generative AI models [generates segment sensitive advertisement assets . . . based on segment dependent characteristics] learned during the machine learning;
[modifying for the user segments] based on the respective generative AI models [a subgroup of the features];
[different asset combinations each of which can be used to display the advertisement for the product] via an online platform.
The additional limitations when viewed individually and when viewed as an ordered combination, and pursuant to the broadest reasonable interpretation, do not integrate the abstract idea into a practical application because each of the additional elements are recited at high level of generality implementing the abstract idea on a computer (i.e. apply it) or generally linking the use of the judicial exception to a particular technological environment. Specifically:
With respect to [receiving] from a data storage, [base advertisement information] and storing, into the data storage [the different image assets], Examiner notes that these limitations are recited at a high level of generality and merely generally link the abstract idea to a particular technological environment (i.e., generic storage to store advertisement information and assets).
With respect to the limitations obtaining, via machine learning, based on the segment-sensitive training data, generative artificial intelligence (Al) models with respect to the user segments, wherein each of the generative AI models with respect to the user segments wherein each of the generative AI models [generates segment sensitive advertisement assets . . . based on segment dependent characteristics] learned during the machine learning; and [modifying for the user segments] based on the respective generative AI models [a subgroup of the features]; Examiner respectfully notes that the use of a machine learning model can be found to integrate an abstract idea into a practical application when the machine learning model is “specifically designed to achieve an improved technology result” (see Ex parte Hannun (Appeal No. 2018-003323) at 10). However, the limitations as claimed are not sufficient to integrate the abstract idea into a practical application. Examiner notes that the limitations obtaining, via machine learning, based on the segment-sensitive training data, generative artificial intelligence (Al) models with respect to the user segments, wherein each of the generative AI models with respect to the user segments wherein each of the generative AI models [generates segment sensitive advertisement assets . . . based on segment dependent characteristics] learned during the machine learning; and [[modifying for the user segments] based on the respective generative AI models [a subgroup of the features]; are recited at a high level of generality. The claims do not specify the type of generative AI algorithms, or how the generative AI models are trained beyond specifying at a high level that machine learning models are obtained based on segment-sensitive training data. Therefore, the trained generative AI models are merely used to generally apply the abstract idea without placing any limits on how the trained generative AI models function. The recitation of claim limitations that attempt to cover any solution to an identified problem with no restriction on how the result is accomplished and no description of the mechanism for accomplishing the result, does not integrate a judicial exception into a practical application or provide significantly more because this type of recitation is equivalent to the words "apply it". See Electric Power Group, LLC v. Alstom, S.A., 830 F.3d 1350, 1356, 119 USPQ2d 1739, 1743-44 (Fed. Cir. 2016); Intellectual Ventures I v. Symantec, 838 F.3d 1307, 1327, 120 USPQ2d 1353, 1366 (Fed. Cir. 2016); Internet Patents Corp. v. Active Network, Inc., 790 F.3d 1343, 1348, 115 USPQ2d 1414, 1417 (Fed. Cir. 2015). Here, Examiner has reviewed Applicant’s specification and notes that the specification does not disclose any particular machine learning or generative AI model. At this level of generality, the recitation of claim limitations that attempt to cover any solution to an identified problem (i.e., obtaining, via machine learning, generic generative AI models to generate advertisement assets with respect to user segments) merely generally links the abstract idea to a technical field/environment, namely a generic computing environment applying machine learning.
Moreover, Examiner notes that Recentive Analytics, Inc. v. Fox Corp. et al., No. 2023-2437, slip op. at 10 (Fed. Cir. Apr. 18, 2025) recently held that claims “that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.” Here, Examiner takes the position that performing the obtaining and modifying limitations using a generic generative AI model is the mere application of generic machine learning to a new data environment. Specifying at a high level what the training data is segment-sensitive training data based on user-segments and corresponding performance metrics is not an improvement to the underlying models. Because no improvement to the underlying machine learning models is disclosed, this limitation does not integrate the abstract idea into a practical application.
Finally, with respect to reciting that the different asset combinations can be used to display the advertisement via an online platform, specifying that the ads can be used can be used to display the ads on an online platform at merely generally links the abstract idea to a particular technological environment (i.e., a generic online platform to display ads) and does not integrate the abstract idea into a practical application.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements when considered both individually and as an ordered combination do not amount to significantly more than the abstract idea. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements are recited at a high level of generality and only generally link the use of the judicial exception to a particular technological environment. The same analysis applies here in 2B, i.e., mere instructions to apply an exception in a particular technological environment cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Alice Corp. also establishes that the same analysis should be used for all categories of claims (e.g., product and process claims). Therefore, machine-readable medium claim 8 and system claim 15 are also rejected as ineligible subject matter under 35 U.S.C. 101 for substantially the same reasons as independent method claim 1. The additional limitations in claim 8 (i.e., a computer-readable medium and machine) and the additional limitations of claim 15 (i.e., a processor, a machine learning engine and AI-assisted asset generator) add nothing of substance to the underlying abstract idea. The components are merely providing a particular technological environment to implement the abstract idea.
Dependent claims 2-7, 9-14, and 16-20 are rejected on a similar rational to the claims upon which they depend. Specifically:
With regard to claims 2-5, 7, 9-12, 14, 16-18, and 20 the additional limitations only serve to further narrow the abstract idea and do not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea.
With regard to claims 6, 13, and, 19, the limitations “analyzing the online feedback information to identify user activities with respect to the previously displayed advertisements; determining the performance metrics of the previously displayed advertisements based on the identified user activities; and detecting the features of advertisement assets used in the previously displayed advertisements” merely additionally narrow the abstract idea. The limitations of “generating the segment-sensitive training data based on the features of advertisements and performance metrics of the previously displayed advertisements; and training the generative AI models using the segment-sensitive training data” are additional elements. However, the limitations as claimed are not sufficient to integrate the abstract idea into a practical application. Specifically, Examiner notes that the generating of training data and training of an AI model based on the training data are recited at a high level of generality and the claims do not specify the type of machine learning algorithm, or how a machine learning model is trained beyond specifying at a high level that features of advertisements and performance metrics of the previously displayed advertisements are utilized. The recitation of claim limitations that attempt to cover any solution to an identified problem with no restriction on how the result is accomplished and no description of the mechanism for accomplishing the result, does not integrate a judicial exception into a practical application or provide significantly more because this type of recitation is equivalent to the words "apply it". See Electric Power Group, LLC v. Alstom, S.A., 830 F.3d 1350, 1356, 119 USPQ2d 1739, 1743-44 (Fed. Cir. 2016); Intellectual Ventures I v. Symantec, 838 F.3d 1307, 1327, 120 USPQ2d 1353, 1366 (Fed. Cir. 2016); Internet Patents Corp. v. Active Network, Inc., 790 F.3d 1343, 1348, 115 USPQ2d 1414, 1417 (Fed. Cir. 2015). Here, Examiner has reviewed Applicant’s specification and notes that the specification does not disclose any particular machine learning or AI model. At this level of generality, the recitation of claim limitations that attempt to cover any solution to an identified problem (i.e., training, via machine learning, generic generative AI models) merely generally links the abstract idea to a technical field/environment, namely a generic computing environment applying machine learning.
Response to Arguments
35 U.S.C. 101
Applicant's arguments, see pages 13-23, filed 2/18/2026 with respect to the rejection(s) of claims
1-20 under 35 U.S.C. 101 have been fully considered but are not persuasive.
First, Applicant argues that:
The Office Action alleges that the claims fall under the "Certain Methods of Organizing Human Activity" grouping of abstract ideas. See, Office Action at page 3. Applicant respectfully disagrees for at least the following reasons. For example, claim 1 is related to machine learning generative AI models based on segment-sensitive training data and utilizing the generative AI models to form different asset combinations used to display an advertisement for a product via an online platform displaying advertisements. Claims 8 and 15 recite similar features. The recited claim features extend far beyond fundamental economic principles or practices (including hedging, insurance, mitigating risk); commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations); managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions), and thus do not fall within the enumerated group of Certain methods of organizing human activity. Also, the above-quoted claim features cannot be performed in the human mind, and thus the claims do not fall within the "mental processes" grouping of abstract ideas. Further, these claim features are not directed to mathematical concepts (remarks page 16).
Examiner respectfully replies that as noted in the 35 U.S.C. 101 rejection above, the claims recite the abstract idea of modifying parts of a base ad asset to create alternate versions for different user segments based on the performance of previously displayed advertisements with respective user segments which is an advertising / marketing activity that falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas. Examiner notes that Examiner does not allege that the claims recite a mental process or mathematical concept and therefore arguments are moot,
Further, Examiner notes that the use of generic machine learning to generate different asset combinations is an additional element that has have been analyzed under step 2A, prong 2, to determine whether it integrates the abstract idea into a practical application. The fact that additional elements are recited in the claim does not negate the fact that the claim recites an abstract idea. Moreover, see Recentive Analytics, Inc. v. Fox Corp. et al., No. 2023-2437, slip op. at 18 (Fed. Cir. Apr. 18, 2025) holding that claims “that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.” Here, the use of generic AI models trained on generic segment-sensitive training data to form different asset combinations is the generic application of machine learning to new data environments without an improvement to the underlying machine learning model.
Second, Applicant argues that:
The claims are patent eligible because the claimed concepts are integrated into a practical application . . . The claims provide an improvement to known technical problems in the technical field of generating electronic content - "[t]raditionally each ad may be assembled and displayed with a particular combination of ad attribute assets for each dynamic-creative optimization (DCO), advertisers need to provide multiple assets for each ad attribute in a predetermined manner regardless of preferences of different locales or individuals" (para. [003 7]). Applicant's claimed concept overcomes such technical problems by "automatically creating ad assets of different attributes based on generative AI" (para. [0037]). "According to the present teaching, an AIassisted ad asset generator may be provided that generates ad assets based on base ad information (e.g., a sports car ad with a base set of assets) in accordance with a generative AI model. The generative AI model may be obtained via machine learning based on feedback information with respect to different display ads previously presented to users of different segments. Such a derived generative AI model may then be applied to each ad to be displayed to create ad assets for different segments with, e.g., respective estimated performance metrics. Such created ad assets with estimated performance metrics may then be used to facilitate, during an ad recommendation process, the selection of a particular asset combination for each DCO ad based on contextual information associated with an ad display opportunity" (para. [0038]). "[T]he trained generative AI model may learn segment-dependent characteristics which may be dynamically adjusted over time based on continually collected feedback information .... Such created ad assets are then stored, at 240, in 160-3 and may be subsequently used for creating different ad asset combinations. When the generative AI model may be trained based on segment-sensitive training data, it may be used to automatically control to generate segment sensitive ad assets with respect to different segments." (Para. [0047]). segments. Such a derived generative AI model may then be applied to each ad to be displayed to create ad assets for different segments with, e.g., respective estimated performance metrics. Such created ad assets with estimated performance metrics may then be used to facilitate, during an ad recommendation process, the selection of a particular asset combination for each DCO ad based on contextual information associated with an ad display opportunity" (para. [0038]). "[T]he trained generative AI model may learn segment-dependent characteristics which may be dynamically adjusted over time based on continually collected feedback information .... Such created ad assets are then stored, at 240, in 160-3 and may be subsequently used for creating different ad asset combinations. When the generative AI model may be trained based on segment-sensitive training data, it may be used to automatically control to generate segment-sensitive ad assets with respect to different segments." (Para. [0047]) (remarks pages 16-19]).
Examiner respectfully disagrees. Examiner has reviewed Applicant’s specification and notes that the specification does not disclose any particular machine learning or generative AI model. At this level of generality, the recitation of claim limitations that attempt to cover any solution to an identified problem (i.e., obtaining, via machine learning, based on the segment-sensitive training data, generative artificial intelligence (Al) models with respect to the user segments, wherein each of the generative AI models with respect to the user segments wherein each of the generative AI models [generates segment sensitive advertisement assets . . . based on segment dependent characteristics] learned during the machine learning; and [modifying for the user segments] based on the respective generative AI models [a subgroup of the features]) merely generally links the abstract idea to a technical field/environment, namely a generic computing environment applying machine learning. Examiner notes that the claimed training and use of generic generative AI models is similar to Claim 2 of Example 47 in the July 2024 Subject Matter Eligibility Examples published by the USPTO available at https://www.uspto.gov/patents/laws/examination-policy/subject-matter-eligibility where the training of a machine learning model using a backpropagation algorithm was not found to integrate the abstract idea into a practical application. Additionally, see Recentive Analytics, Inc. v. Fox Corp. et al., No. 2023-2437, slip op. at 18 (Fed. Cir. Apr. 18, 2025) holding that claims “that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.” Here, Examiner takes the position that performing the obtaining and modifying limitations using a generic generative AI model is the mere application of generic machine learning to a new data environment. Specifying at a high level what the training data is “segment-sensitive” is not an improvement to the underlying models. At most, the claims recite the use of generic machine learning models obtained via generic segment-sensitive training data with no technical explanation of how the training occurs. Because no improvement to the underlying machine learning models is disclosed, this limitation does not integrate the abstract idea into a practical application.
Third, Applicant argues that:
Claim 1 provides improvement to the machine-learned generative AI models by using segment-sensitive training data so that the machine-learned generative AI models learn segment-dependent characteristics which may be dynamically adjusted over time based on continually collected feedback information and automatically control to generate segment-sensitive ad assets with respect to different user segments (remarks page 19).
Examiner respectfully disagrees and notes that the claims merely recite obtaining machine learning models based on segment sensitive training data. There is no recitation of continually collecting feedback information and dynamically adjusting segment dependent characteristics based on the continually collected feedback.
Fourth, Applicant argues that the generating and modifying limitations are:
not recited at a high level of generality and are not mere generic computing environment or mere application of generic machine learning to a new data environment, at least because the claims recite details of the machine learning - the machine learning of the generative AI models being based on segment-sensitive training data generated based on online feedback information on previously displayed advertisements, and the details of modifying - the modifying being with respect to at least one of the features exhibited in the base image to create segment-sensitive advertisement assets of the product (remarks page 20).
Examiner respectfully disagrees and replies that the claims do not recite technical details of the machine learning. Examiner notes that unlike the claims in Ex Parte Desjardins (which recited an improved way of training a machine learning model to protect the model’s knowledge about previous tasks), the present claims do not recite a technical improvement. Specifically, merely specifying at a high level of generality that the training data is segment-sensitive training data generated based on feedback and that generative AI models are used to create segment-sensitive ad assets of the products are not technical details that amount to an improved way of training a machine learning model. Because no improvement to the underlying machine learning models is disclosed, this limitation does not integrate the abstract idea into a practical application.
Fifth, Applicant argues:
In addition to failing to consider Applicant's claims as an ordered combination and as a whole, the Office Action has improperly analyzed the claims without considering the "additional element(s)" in combination with the non-additional elements. As a result, the Office Action has also incorrectly and improperly identified that the alleged "additional elements" do not amount to significantly more than the alleged judicial exception (remarks page 22).
Examiner respectfully disagrees that the Office action has improperly analyzed the claims without considering the additional elements in combination the non-additional elements. Specifically, Examiner notes that that all limitations of the claims have been considered both individually and as an ordered combination, and the additional elements have not been found to integrate the abstract idea into a practical application because each of the additional elements are recited at high level of generality implementing the abstract idea on a computer (i.e. apply it) or generally linking the use of the judicial exception to a particular technological environment.
Finally, Applicant argues that:
Further, the Office Action on pages 24 and 25 alleges that "Examiner ... notes that the specification does not disclose any particular machine learning or generative AI model." Applicant respectfully disagrees and submits that at least paras. [0038], [0040], [0045], [0047], [0053], [0054]-[0066] (written description of FIGS. 5A-5D that describes how the generative AI models are used to control asset creation process, written description of FIG. 6A that describes how the AI models are obtained, written description of FIG. 6B that describes how the AI models are used to generate assets for an advertisement) describe the generative AI model obtained via machine learning based on segment-sensitive training data (remarks pages 22-23).
Examiner respectfully disagrees and maintains that the specification does not disclose the specific model or the specify way the model is trained beyond specifying the training data at a high level of generality (e.g., a generic generative AI model trained based on segment-sensitive training data – see spec paragraph [0047]). The obtaining and modifying limitations using a generic generative AI model is the mere application of generic machine learning to a new data environment. Specifying at a high level what the training data is generated from is not an improvement to the underlying models. Because no improvement to the underlying machine learning models is disclosed, this limitation does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea.
35 U.S.C. 103
Applicant's arguments, see pages 23-25, filed 02/18/2026, with respect to the rejections of claims 1-20 under 35 U.S.C. 103 have been fully considered and are persuasive. Therefore the rejections have been withdrawn.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US Patent Application Publication Number 20190392487 (“Duke”) discloses training an algorithm to assemble ads using the optimal combination of ad-components and layout
US Patent Application Publication Number 20250069298 (“Srinivasan”) discloses generation of synthetic images of products and product variations using a fine-tuned diffusion model
US Patent Publication Number 11113715 (“Schmutz”) discloses generating a plurality of creative content permutations corresponding to creative content using a plurality of image and textual elements and parameter settings from among variable parameters wherein the creative content permutations have respective selection probabilities
US Patent Application Publication Number 20230316342 (“Sebag”) discloses creating permutations of a single digital advertisement by combining creative assets extracted from the single digital advertisement with many different presentation layers to form a multitude of digital advertisement and presentation layer pairs
US Patent Application Publication Number 20190095959 (“Yu”) discloses using generative models to general advertisement components
US Patent Application Publication Number 20240029103 (“Myers”) discloses creating new ad variations with precise and relevant targeting using generative adversarial networks
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 ALLAN J WOODWORTH, II whose telephone number is (571)272-6904. The examiner can normally be reached Mon-Fri 9:00-5:30.
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, Ilana Spar can be reached at (571) 270-7537. 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.
/ALLAN J WOODWORTH, II/
Primary Examiner, Art Unit 3622
Prosecution Insights