DETAILED ACTION
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
2. The following office action is a Final Office Action in response to the communications received on 03/02/2026.
Claims 1, 2, 7, 9, 11-13 and 18-20 have been amended. Therefore, claims 1-20 are currently pending in this application.
Response to Amendment
3. The amendment to each of claims 1, 11 and 20 is sufficient to overcome the rejection set forth in the previous office action under section §112(a). Accordingly, the Office withdraws the above rejection.
Claim Rejections - 35 USC § 101
4. Non-Statutory (Directed to a Judicial Exception without an Inventive Concept/Significantly More)
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.
(Step 1)
The current claims fall within one of the four statutory categories of invention (MPEP 2106.03).
Step 2A [Wingdings font/0xE0] Prong One:
The claim(s) recite a judicial exception, namely an abstract idea, as shown below:
— Considering each of claims 1, 11 and 20 as representative claims, the following claimed limitations recite an abstract idea:
Claim 1:
present a prompt assignment comprising instructions for teaching a user how to construct prompts [for] conversational exchanges;
receive an initial user-constructed prompt in response to the prompt assignment;
[submit] the initial user-constructed response;
receive one of more prompt suggestions to guide revision of the initial user-constructed prompt, the one or more prompt suggestions [created] by classification of the initial user-constructed prompt according to one or more predefined prompt characteristics;
[present] the one or more prompt suggestions in accordance with the initial user-constructed prompt;
receive a revised prompt based on the one or more prompt suggestions;
submit the revised prompt;
receive a reply to the revised prompt;
submit the revised prompt and the reply;
receive a prompt insight based on classification of the revised prompt and the reply according to one or more prompt characteristics;
[present] the reply together with the prompt insight in the context of the prompt assignment.
Claim 11 and 20:
provide a prompt assignment comprising instruction for teaching a user how to construct prompts;
receive first an initial user-constructed prompt in response to the prompt assignment;
[submit] the initial user-constructed prompt;
receive one of more prompt suggestions to guide revision of the initial user-constructed prompt, the one or more prompt suggestions [created] by classification of the initial user-constructed prompt according to one or more predefined prompt characteristics;
[present] the one or more prompt suggestions in accordance with the initial user-constructed prompt;
receive a revised prompt based on the one or more prompt suggestions;
submit the revised prompt;
receive a first reply to the revised prompt;
submit the revised prompt and the first reply;
receive a prompt insight based on classification of the revised prompt and the first reply according to one or more prompt characteristics;
provide the first reply and the prompt insight.
Thus, the limitations identified above recite an abstract idea since the limitations correspond to certain methods of organizing human activity, and/or mental processes, which are part of the enumerated groupings of abstract ideas identified according to the current eligibility standard (see MPEP 2106.04(a)).
For instance, the current claims correspond to managing personal behavior (e.g., teaching), wherein a user is presented with a prompt assignment, which comprises instructions for teaching a user how to construct prompt for conversational exchanges. Thus, based on a user-constructed prompt, which is received from the user in response to the prompt assignment, one or more prompt suggestions for guiding the revision of the user-constructed prompt are generated based on classifying the user-constructed prompt according to one more predefined prompt characteristics; and furthermore, after presenting one or more of the above suggestions to the user, a revised prompt is received from the user and submitted; and in response, the user receives a reply to the revised prompt; and accordingly, after the user submits the revised prompt and the above reply, the user is presented with the reply and the prompt insight in the context of the assignment, wherein the prompt insight is obtained based on classifying the revised prompt and the reply according one or more predefined prompt characteristics, etc.
Similarly, given the fact that the core of the claimed process can be performed in the human mind (and/or using a pen and paper), the claims also correspond to the group mental processes—i.e., observation, evaluation, judgment, opinion, etc. For instance, considering current claim 1 as an example, a human—such as a teacher—can perform—mentally and/or using a pen and paper—the crux of the claimed process as follows:
the teacher presents, using a pen and paper, a prompt assignment to the user; wherein the prompt assignment comprises instructions for teaching the user how to construct prompts (e.g., how to construct a prompt, which the user submits to an entity for conversational exchanges, etc.);
the teacher then receives, using a pen and paper, an initial user-constructed prompt from the user in response to the assignment above;
the teacher evaluates the initial user-constructed response by consulting one or more entities—such as, one or more publications and/or guidebooks related to constructing prompts, etc.;
the teacher then drafts, using a pen and paper, one or more prompt suggestions to guide revision of the initial user-constructed prompt; the teacher drafts the one or more prompt suggestions above based on classifying the initial user-constructed prompt according to one or more predefined prompt chrematistics;
the teacher presents to the user, using a pen and paper, the one or more prompt suggestions above in accordance with the initial user-constructed prompt;
the teacher then receives from the user a revised prompt, which the user made based on the one or more prompt suggestions;
the teacher then consults one or more of the entities (i.e., one or more of the publications and/or guidebooks) to evaluate the revised prompt received from the user, including: (i) drafting a reply to the revised prompt, (ii) analyzing the revised prompt and the reply, and (iii) drafts a prompt insight based on classifying the revised prompt and the reply according to one or more predefined prompt characteristics;
the teacher then presents, using a pen and paper, the reply together with the prompt insight in the context of the prompt assignment.
The analysis above confirms that the current claims do recite an abstract idea; namely, a mental process.
Step 2A [Wingdings font/0xE0] Prong Two:
The claims recite additional element(s), wherein a computer device/system with basic components (e.g., a processor, a storage media, etc.), which executes an algorithm—such as, a generative artificial intelligence model, is utilized to facilitate the recited functions/steps regarding: displaying information to a user (e.g., “present, via a learning platform, a prompt assignment comprising instructions for teaching a user how to construct prompts for submission to a generative artificial intelligence (AI) model, wherein the generative AI model supports conversational exchanges with the user”); collecting a response(s)/input(s) from the user (e.g., “receive, via a user interface of the learning platform, user input comprising an initial user-constructed prompt in response to the prompt assignment”); transmitting, to a service provider, data gathered from the user (e.g., “transmit the initial user-constructed prompt to an insights service integrated with the learning platform”); collecting further data (e.g., “receive, at the learning platform, one or more prompt suggestions configured to guide revision of the initial user-constructed prompt, the one or more prompt suggestions generated by the insights service based on automated classification of the initial user-constructed prompt according to one or more predefined prompt characteristics”); displaying information to the user (e.g., “display, via the user interface of the learning platform, the one or more prompt suggestions in accordance with the initial user-constructed prompt”); collecting further input from the user (e.g., “receive, via the user interface, a revised prompt based on the one or more prompt suggestions”); submitting the received input to a model (e.g., “submit the revised prompt to the generative AI model”); receiving a reply from the model (e.g., “receive, from the generative AI model, a reply to the revised prompt”); submitting further information to the service provider (e.g., “submit, by the learning platform, the revised prompt and the reply generated by the generative AI model to the insights service”); collecting a result(s) from the service provider (e.g., “receive, at the learning platform, a prompt insight based on automated classification of the revised prompt and the reply according to one or more predefined prompt characteristics”); displaying to the user the result(s) obtained from the service provider (e.g., “display the reply together with the prompt insight in the context of the prompt assignment”), etc.
However, the claimed additional element(s) fail to integrate the abstract idea into a practical application since the additional element(s) are utilized merely as a tool to facilitate the abstract idea. Thus, when each claim is considered as a whole, the additional element(s) fail to integrate the abstract idea into a practical application since they fail to impose meaningful limits on practicing the abstract idea. Although the claims utilize an algorithm—namely, a generative artificial intelligence (AI) model—to generate pertinent information based on the analysis of input data collected from the user (e.g., input in the form of a user-constructed prompt, etc.), neither the current claims nor the original disclosure signifies any technological improvement. Instead, each of the current claims, including the original disclosure, is utilizing the existing computer/network technology—merely as a tool—to facilitate the presentation of pertinent information to the user, based on the analysis of input data collected from the user, etc.
Accordingly, when each of the claims is considered as a whole, none of the claims provides an improvement over the relevant existing technology.
The observations above confirm that the claims are indeed directed to an abstract idea.
Step 2B
Accordingly, when the claim(s) is considered as a whole (i.e., considering all claim elements both individually and in combination), the claimed additional elements do not provide meaningful limitations to transform the abstract idea into a patent-eligible application of the abstract idea such that the claim(s) amounts to “significantly more” than the abstract idea itself (also see MPEP 2106). The claimed additional elements are directed to conventional computer elements, which are serving merely to perform conventional computer functions.
Accordingly, when each of the current claims is considered as a whole (e.g., see the discussion under Prong Two above regarding such consideration of the claim as a whole), none of the claims recites an element—or a combination of elements—directed to an inventive concept.
It is also worth to note that the utilization of the conventional computer/network technology to facilitate the presentation of pertinent information to a user, including facilitating a teaching/learning process, wherein the user is presented with one or more pertinent feedback based on the analysis of the user’s response(s) to one or more assignments, etc., is directed to a well-understood, routine, conventional activity in the art (e.g., US 2016/0343272; US 2017/0091312; US 2012/0329029; US 2009/0198488; US 2012/0034591, etc.).
Note also that the use of a generative artificial intelligence, as an interactive tool, to facilitate human-machine interaction, including engaging a user with a more realistic human-like natural conversations, etc., is directed to a well-understood, routine, conventional activity in the art (e.g., see US 2018/0020093; US 2018/0376002; also see “Deep Reinforcement Learning For Dialog Generation”, the Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1192–1202, Austin, Texas, November 1-5, 2016. ©2016 Association for Computational Linguistics.
The observations above confirm that the current claimed invention fails to amount to “significantly more” than an abstract idea.
It is worth noting that the above analysis already encompasses each of the current dependent claims (i.e., claims 2-10 and 12-19). Particularly, each of the dependent claims also fails to amount to “significantly more” than the abstract idea since each dependent claim is directed to a further abstract idea, and/or a further conventional computer element(s) utilized to facilitate the abstract idea.
Accordingly, the findings above demonstrate that none of the claims implements an element—or a combination of elements—directed to an inventive concept (e.g., none of the current claims is reciting an element—or a combination of elements—that provides a technological improvement over the existing/conventional technology).
► Applicant’s arguments directed to section §101 have been fully considered (the arguments filed on 06/11/2026). However, the arguments are not persuasive at least for the following reasons:
Firstly, while citing several paragraphs from the specification, which are assumed to describe the various features that the current claims are reciting (see pages 12-13 of Applicant’s arguments), Applicant asserts that “[t]hese amended features collectively define a multi-stage, feedback-driven processing pipeline that improves the operation of the computer system itself, rather than merely improving user understanding or behavior. The system transforms unstructured prompting interactions into structured classification data that is used to control downstream processing, thereby enabling the system to operate on prompting data in a repeatable and machine-interpretable manner. The use of automated classification of prompting activity data at multiple stages introduces a structured data model that reorganizes interaction data into defined categories such as conciseness, clarity, focus, or ambiguity, allowing the system to process and act on prompt quality in a way that is not available to human users alone. This transformation of data constitutes an improvement in computer functionality analogous to the type of structural data improvement recognized in Enfish, LLC v. Microsoft Corp., 822 F.3d 1327 (Fed. Cir. 2016) (‘Enfish’) and expressly endorsed in the USPTO's memo addressing Ex parte Desjardins (Appeal No. 2024-000567) entitled Advance notice of change to the MPEP in light of Ex Parte Desjardins (‘Desjardins Memo’)” (emphasis added).
However, except for the attempt made to show the support, which the specification is providing regarding the claimed features, Applicant fails to demonstrate an element (if any)—or a combination of elements (if any)—that supposedly provides the alleged technological improvement. For instance, except for the conclusory assertion regarding the alleged “multi-stage, feedback-driven processing pipeline”, which is assumed to “improve[] the operation of the computer itself”, Applicant fails to show what feature (if any) of the computer system is being improved. Similarly, Applicant’s generic assertion regarding the alleged transformation of “unstructured prompting interactions into structured classification data that is used to control downstream processing”, which supposedly enables the system to “operate on prompting data in a repeatable and machine-interpretable manner”, also does not show whether any of the current claims (or even the discloser) is implementing an element—or a combination of elements—that provides a technological improvement. This is because the process of gathering raw (unstructured) data from a source (e.g., a task where a cashier randomly scans various products that a customer is purchasing from a store) and automatically generating structured data based on analyzing the raw data (e.g., generating a receipt/table, which lists the various products according their classifications—such as: electronics, grocery, clothes, etc.), is already part of the existing computer/network technology. Thus, the mere use of such features of the existing computer technology in a particular field (e.g., the online learning field, the medical field, etc.) does not constitute a technological improvement. Accordingly, Applicant’s attempt to emphasize the benefits that such existing feature is providing to the claimed system/method—namely, the alleged introduction of “structured data model that reorganizes interaction data into defined categories such as conciseness, clarity, focus, or ambiguity”, also fails to demonstrate a technological improvement (if any) that the claimed—or the disclosed—system/method is providing. In particular, the benefits, which Applicant is alleging regarding converting unstructured data into structured data, are inherent attributes of the existing computer/network technology.
Moreover, unlike Applicant’s assertion, neither Enfish nor Ex Parte Desjardins is analogous to any of the current claims. For instance, Enfish provides a technological improvement in terms of: memory management (e.g., a smaller storage space requirement), faster search times, etc. In contrast, except for using the existing computer/network technology—merely as a tool—to facilitate learning, neither the current clams nor the original disclosure provides any technological improvement. Similarly, Desjardins provides a technological improvement at least in terms an advanced technique for training a machine-learning model, which allows the model to learn new tasks while preserving performance on earlier tasks; and this implementation addresses the technical problem of “catastrophic forgetting” in continual learning systems. In contrast, neither the current claims nor the original disclosure even contemplated—much less actually implemented—any advanced technique for training a machine-learning or AI model. Instead, even the specification admits that the disclosed system/method is relying merely on existing training techniques, including: supervised learning, unsupervised learning, etc. (e.g., see [0022] to [0024] of the original specification).
The observations above confirm that neither Enfish nor Desjardins is analogous to any of the current claims; and consequently, Applicant’s arguments are not persuasive.
Note also that Applicant’s attempt to compare the computer-based process to a human is also not relevant to show a technological improvement (if any) that Applicant is alleging. This is again because a technological improvement (if any) is demonstrated by comparing two technologies—such as, the existing computer/network technology against the technology that Applicant is claiming/disclosing. Consequently, Applicant’s argument in this regard is also not persuasive.
In addition, while referring to various paragraphs from the specification, Applicant is asserting that “the amended claim recites a distributed, role-based architecture in which the learning platform, insights service, and generative AI model perform distinct and coordinated functions . . . The insights service operates as a dedicated analysis engine that performs automated classification and generates suggestions, while the generative AI model operates as an execution engine that produces responses, and the learning platform orchestrates the interaction between these components . . . This computational structure enables modular processing and scalable operation, which is a technical system improvement . . . As emphasized in Ex parte Carmody (Appeal No. 2025-002843), claims reciting a specific arrangement of multiple models and components operating in an ordered combination may constitute a patent-eligible improvement in system functionality rather than an abstract idea” (emphasis added).
However, except for relying on yet another subjective assumption to substantiate the alleged “technical system improvement”, which the alleged “distributed, role-based architecture” is supposedly providing, Applicant still fails to identify any claimed (or disclosed) feature—or any combination of claimed/disclosed features—that provides a technological improvement over the relevant existing technology. In contrast, even the existing Internet technology, which implements a vast client-server architecture, already coordinates multiple components that interact with one another, including a distributive computing scheme in which tasks are shared among various components. Accordingly, Applicant fails to demonstrate whether the alleged architecture, which supposedly coordinates “the learning platform, insights service, and generative AI model”, is beyond the existing Internet technology. Instead, Applicant appears to simply summarize the functions that the above components are assumed to be performing individually and/or in coordination.
Moreover, unlike Applicant’s assertion, neither the current claims nor the original disclosure implements any “specific arrangement of multiple models and components”, which supposedly provides the alleged technological improvement. In particular, as already pointed out above, the claimed and the disclosed arrangement corresponds to the existing client-server arrangement, which is again one of the fundamental features of the existing Internet technology. Consequently, Applicant’s attempt to justify an alleged technological improvement, while misapplying the Board’s decision regarding Ex parte Carmody (Appeal No. 2025-002843), is certainly not persuasive. It is worth noting that Carmody is considered to provide a technological improvement since it implements a specific scheme for training a machine-learning model. In particular, as already indicated per the Board’s analysis, Carmody implements the modular approach or the plug-and-play module, which “enables the model for each tactic to be updated and improved separately and independently from other tactic-specific models, and also enables models for new tactics to be easily incorporated into tactic recommendation model” (see page 7 of Appeal 2025-002843, emphasis added). Thus, the Board identified the above as the technological improvement described per Carmody’s specification. In contrast, as already pointed out above, nether the current claims nor the original disclosure teaches any advanced scheme for training any ML or AI model. Instead, the original disclosure is relying merely on existing training techniques (see above the parts cited from the specification).
Appellant further asserts, “the claim recites a feedback-driven control loop in which prompting activity data is repeatedly analyzed, refined, and reprocessed . . . system does not analyze a prompt once, but it captures initial prompt data, generates suggestions, processes a revised prompt through the AI model, and then reanalyzes the resulting prompt-reply pair to generate further insights. This closed-loop architecture allows the system to dynamically adjust its processing of prompts based on observed outcomes, thereby improving how the system operates over time . . . Such a feedback mechanism constitutes a technical control process that modifies system behavior . . . The architecture also functions as a pre-processing gateway for inputs to the generative AI model. By generating classification-based suggestions before the revised prompt is submitted to the model, the system conditions input data into a more effective form prior to model execution. This results in more efficient downstream processing by reducing the need for repeated prompt submissions and iterative reprocessing of prompt data, and supports a system in which prompt quality is improved prior to submission and interactions are iteratively refined . . . There is a direct technical result that fewer redundant processing cycles are required, which constitutes an improvement in how the system manages computational workflows and processes input data” (emphasis added).
However, except for making several conclusory assertions regarding the alleged improvements, Applicant still fails to address the crux of the eligibility analysis; namely, demonstrating an element (if any)—or a combination of elements (if any)—that provides a technological improvement over the relevant existing technology. For instance, Applicant’s alleged “feedback-driven control loop” does not constitute a technological improvement regardless of whether it is considered individually or in combination with the rest of the clamed or disclosed features. This is because the implementation of chatbots, which execute one or more ML and/or AI models that analyze textual and/or audio data, is already part of the existing computer/network technology. Accordingly, regardless of whether the claimed (and/or the disclosed) system is analyzing, based on executing an AI model, the user’s input/prompt only once or several times, the implementation still does not constitute a technological improvement. In particular, depending on whether the received prompt is ambiguous or clear, existing chatbots typically adapt their response. For instance, if the user’s initial prompt is ambiguous or unclear, a chatbot may execute one or more dialogue steps—such as, a dialog step(s) that requests the user to provide further clarification, and/or a dialog step(s) that requires the user to choose the most relevant prompt from a set of alternatives, etc. In contrast, despite relying on such features of the existing computer/network technology, Applicant is attempting to show an alleged technological improvement that the current claims are assumed to be providing. Consequently, none of Applicant’s conclusory assertions (e.g., the alleged improvement regarding “how the system operates over time” and/or the alleged “technical control process that modifies system behavior”, etc.) is persuasive.
Of course, the same is true regarding Applicant’s alleged “more efficient downstream processing”, which supposedly reduces “the need for repeated prompt submissions and iterative reprocessing of prompt data”. In particular, while repeatedly relying on the features of the existing computer/network technology, Applicant is once again attempting substantiate the alleged technological improvement that the current claims are assumed to be providing. For instance, similar to the point made above, existing chatbots normally perform one or more desired tasks (e.g., providing relevant information or answer to the user) once the chatbots properly construe the user’s request. Accordingly, if the user’s prompt (request) is unclear or ambiguous, the chatbot normally provides one or more dialog steps in order to accurately construe the user’s request (e.g., requesting the user to provide clarification; requesting the user to choose the most relevant prompt from a set of alternatives, etc.). In this regard, depending on the skill level of the user, the above dialogue may have just few steps or several steps before the chatbot properly construes the user’s request. Of course, once the chatbot properly construes the user’s request, it proceeds with retrieving the relevant information or answer to the user, etc.
In contrast, despite relying on such features of the existing computer/network technology, Applicant is making various conclusory assertions regarding some alleged technological improvements; namely, the system’s characteristic to: (a) condition “input data into a more effective form prior to model execution”, (b) achieve “more efficient downstream processing”, (c) improve “prompt quality”, (d) reduce “redundant processing cycles”, (e) improve “how the system manages computational workflows and processes input data”, etc. Consequently, Applicant’s arguments are not persuasive. If anything, Applicant appears to be mistaking some of the inherent attributes of the existing computer/network technology for a technological improvement.
Secondly, while referring to Prong One of Step 2A, Applicant asserts, “[w]hen properly analyzed, amended claim 1 is not directed to a judicial exception. The claim does not merely recite a mental process or organizing human activity, but instead recites a machine-implemented workflow involving automated classification and inter-service data processing that cannot practically be performed in the human mind . . . depend on system-level data structures, inter-component communication, and persistent data tracking, all of which are beyond the capacity of a human mind . . . the claim is not directed to organizing human activity merely because it involves a ‘prompt assignment’. The focus of the claim is not on teaching or evaluating user behavior, but on how the computer system processes prompting data through a defined pipeline involving multiple specialized components. As discussed in Ex parte Desjardins, the inquiry focuses on whether the claims are directed to an improvement in computer functionality . . . the claim is directed to a system-level data processing architecture that improves how prompting interactions are handled within the computing system” (emphasis added).
However, quite similar to the point made in the previous office action, Applicant once again fails to properly apply the inquiry under Prong One of Step 2A. In particular, while emphasizing the claimed computer elements, Applicant is attempting to challenge the Office’s findings presented under Prong One. In contrast, Prong One of Step 2A does not require one to consider any of the computer element, which are part of the additional elements. Instead, while excluding the computer elements (e.g., the alleged “machine-implemented workflow”, “inter-service data processing”, etc.), Prong One requires one to identify merely the limitations that recite the judicial exception (see MPEP 2106.07(a)). Consequently, unlike Applicant’s new theory, neither Desjardins nor the MPEP required Prong One to focus on an improvement in computer functionality.
Regarding Prong Two of Step 2A, Applicant asserts that “[t]he claim recites a specific technical solution that improves how prompting interactions are processed within a computing system by implementing an ordered combination of operations involving automated classification, structured data processing, and coordinated inter-component communication . . . examiners should evaluate whether the claims reflect an improvement described in the specification and must consider the claim as a whole, rather than at a high level of generality. The Desjardin Memo further emphasizes that claims do not need to explicitly state the improvement, but must include the steps or components that implement it . . . claim 1, analyzed as a whole, includes automated classification of prompting activity data and iterative generation of suggestions and insights based on that classification” (emphasis added).
However, once again Applicant appears to rely on conclusory assertions, as opposed to factual findings, to substantiate the alleged technological improvement. For instance, Applicant’s alleged “specific technical solution”, which supposedly “improves how prompting interactions are processed within a computing system”, does not signify a particular feature (if any)—or a combination of features (if any)—that constitutes a technological improvement. Of course, due to the lack of such a feature(s), Applicant is asserting that the alleged “specific technical solution” is the result of “an ordered combination of operations involving automated classification, structured data processing, and coordinated inter-component communication”. However, such generic assertion also fails to point out the feature (if any)—or the combination of features (if any)—that supposedly provides the technological improvement that Applicant is alleging. For instance, the process of implementing one or more algorithms that automatically classify, based on one or more attributes, collected data elements into one or more categories, etc., is already one of the fundamental features of the existing computer and/or network technology. In fact, besides structuring collected information, the existing Internet technology incorporates one or more search engines, which automatically classify data elements (e.g., search queries, documents, etc.) into one or more relevant categories before and/or during the execution of search tasks. Of course, basic common sense also dictates that the existing Internet technology, which is part of the exiting computer/network technology, already coordinates various inter-component communications.
The observation above demonstrates that Applicant is attempting to show a technological improvement while essentially signifying the features of the existing Internet technology. Thus, regardless of whether the features are being considered individually or in ordered combination, none of the features that Applicant is alleging, including the alleged “automated classification of prompting activity data” and/or the alleged “iterative generation of suggestions and insights based on that classification”, is directed to a technological improvement.
In addition, the Office’s analysis already considers the description in the specification. In fact, the analysis already confirms the lack of description regarding any feature—or any combination of features—directed to a technological improvement. For instance, regarding the AI that Applicant is repeatedly emphasizing, the Office’s analysis points out that the specification does not even implement a particular manner of training the AI, much less a training technique considered to be an advance over the existing computer technology. Instead, the specification is merely providing a generic description regarding the use of existing schemes (see [0022] to [0024]). Accordingly, the Office’s analysis already complies with the Desjardins memo. In contrast, despite the complete lack of description in the specification regarding any technological improvement, Applicant repeatedly emphasizes the Desjardins memo as if the Office overlooked a part of the description that describes a technological improvement. Thus, Applicant’s arguments are not relevant. In particular, given the complete lack of description regarding any technological improvement, the Office is not expected to find a nonexistent technological improvement.
Applicant further asserts, “[c]laim 1 also reflects the type of specific, ordered architecture that was found eligible in Ex parte Carmody, where the Board emphasized that a particular arrangement of models and processing steps can integrate an abstract idea into a practical application. Like the claims in Ex parte Carmody, amended claim 1 recites a defined architecture in which distinct components perform specific roles in a coordinated processing sequence . . . claim 1 addresses a technical problem in prompting systems. For example, claims 1 includes process that refine prompt inputs in a structured and repeatable manner by implementing a classification-driven feedback loop that conditions input data and enables iterative refinement. This constitutes a particular solution implemented through specific technical means, rather than a generalized instruction to analyze or display information” (emphasis added).
However, Applicant is attempting to substantiate the alleged technological improvement while misapplying the Board’s decision regarding Carmody (i.e., Appeal 2025-002843). In particular, unlike Applicant’s theory, Carmody is patent-eligible not simply because it has a particular arrangement of models and processing steps, but because it that an arrangement that provides a technological improvement. In particular, as already pointed out above, Carmody implements a particular scheme for training a machine-learning model. More specifically, Carmody implements the modular approach or the plug-and-play module, which “enables the model for each tactic to be updated and improved separately and independently from other tactic-specific models, and also enables models for new tactics to be easily incorporated into tactic recommendation model” (again see page 7 of Appeal 2025-002843, emphasis added). Of course, the Board identified the above as the technological improvement described per Carmody’s specification.
In contrast, neither Applicant’s current claims nor the original disclosure as a whole provides any feature—or any combination of features—directed to a technological improvement. Even when considering the claimed/disclosed AI model, the original specification does not even contemplate—much less actually implement—any advanced scheme for training any ML or AI model. Instead, the original disclosure is relying merely on existing training techniques (again [0022] to [0024] of the specification). Consequently, Applicant’s attempt to substantiate an alleged technological improvement, while misapplying the Board’s decision regarding Carmody, is once again not persuasive.
Similarly, Applicant’s generic assertions also fail to demonstrate whether any of the current claims, or even the specification, is implementing a technological improvement. For instance, Applicant’s alleged “technical problem”, which is assumed to relate to “prompting systems”, does not appear to be logical. For instance, simply labeling the existing computer system as a “prompting system” does not necessarily imply a technological distinction. Moreover, Applicant’s alleged solution to the alleged “technical problem” is a process that “refine[s] prompt inputs in a structured and repeatable manner by implementing a classification-driven feedback loop that conditions input data and enables iterative refinement”; and accordingly, Applicant is essentially describing the processes that existing chatbots are performing (e.g., see the discussion above in this regard). The above confirms that Applicant is attempting to substantiate the alleged technological improvement while mistaking the features of the existing technology for a technological improvement. Consequently, none of Applicant’s conclusory assertions, including the alleged “particular solution implemented through specific technical means”, is persuasive.
Thirdly, regarding Step 2B, Applicant is asserting that “[c]laim 1 recites significantly more than a judicial exception . . . The learning platform, insights service, and generative AI model are not recited as generic components performing generic functions. Instead, they operate in a coordinated manner in which prompting activity data is transmitted, classified, used to generate suggestions, and reprocessed in a feedback loop. This ordered combination of steps including pre-processing of prompt data, classification-driven suggestion generation, model execution, and post-processing of prompt and reply data, constitutes a technological solution rather than routine or conventional activity” (emphasis added).
However, except for the attempt made to summarize the functions that the claimed features are assumed to perform in coordinated manner, Applicant fails to demonstrate an arrangement (if any) that is considered to be beyond the conventional computer/network technology. For instance, Applicant fails to point out a feature (if any), or a combination of features (if any), that makes any one or more of the components that Applicant listed (i.e., the learning platform, the insights service, the generative AI model, etc.) beyond the conventional technology. In contrast, even the original specification admits that the generative AI model, which Applicant is repeatedly emphasizing, does not implement any new or advanced feature; rather, it is directed to the existing technology (see [0022] to [0024]). In fact, per the original specification, the disclosed system as a whole is directed to the conventional computer/network technology (e.g., see [0031], [0032], [0034], [0035], emphasis added),
“. . . Operational environment 100 includes learning platform 101, foundation model service 105, and insights service 111, as well as computing devices 120, 130, and 140. Learning platform 101 employs one or more server computers 103 co-located with respect to each other or distributed across one or more data centers. Example servers include web servers, application servers, virtual or physical servers, or any combination or variation thereof . . .”
“ Computing devices 120, 130, and 140 communicate with learning platform 101 via one or more internets and intranets, the Internet, wired and wireless networks . . . Examples of computing devices 120, 130, and 140 include personal computers, tablet computers, mobile phones . . .”
“ Learning platform 101 includes an integration (e.g., an application programming interface) with foundation model service 105 to support conversational interactions between observed users and foundation model-powered chatbots and other types of AI tools. Foundation model service 105 employs one or more server computers . . .”
“ Learning platform 101 also includes an integration with insights service 111, which is capable of analyzing and reporting on the conversational interactions between observed users and foundation model engine 109. Insights service 111 employs one or more server computers . . .”
The excerpts above confirm that the disclosed system as a whole is directed to the conventional client-server arrangement, which is indeed the conventional computer/network technology. Of course, such finding further confirms the fact that each of the current claims, when considered as a whole, is directed to the conventional and generic arrangement of the additional elements. Consequently, none of the ordered combination of the current claimed features, including Applicant’s alleged ordered combination of “pre-processing of prompt data, classification-driven suggestion generation, model execution, and post-processing of prompt and reply data”, constitutes an arrangement beyond the conventional computer/network technology. Consequently, Applicant’s conclusion regarding the alleged “technological solution” is not persuasive.
Applicant further asserts, “[a]s discussed in Ex parte Desjardins and reinforced in the Desjardins memo, claims that reflect a specific improvement in system operation based on an ordered combination of steps are not merely conventional, even if they involve known components . . . Ex parte Carmody confirms that a particular architecture involving model-based processing can constitute an inventive concept . . . The Office Action has not established that this specific classification-driven, multi-stage processing architecture was well-understood, routine, or conventional” (emphasis added).
However, except for presenting a very broad and generalized assumption regarding the two recent decisions, Desjardins and Carmody, Applicant still fails to demonstrate whether any of the above decisions is relevant to any of the current claims. In contrast, as already pointed out above, neither Desjardins nor Carmody is relevant to any of the current claims. In particular, unlike the current claims and the original specification, each of Desjardins and Carmody implements a technological improvement. For instance, Desjardins provides a technological improvement at least in terms an advanced technique for training a machine-learning model, which allows the model to learn new tasks while preserving performance on earlier tasks; and this implementation addresses the technical problem of “catastrophic forgetting” in continual learning systems. Similarly, Carmody also provides a technological improvement since it implements a specific scheme for training a machine-learning model—such as, the modular approach or the plug-and-play module, which “enables the model for each tactic to be updated and improved separately and independently from other tactic-specific models, and also enables models for new tactics to be easily incorporated into tactic recommendation model” (again see page 7 of Appeal 2025-002843, emphasis added).
Thus, each of Desjardins and Carmody involves a concrete implementation that signifies a technological improvement at least in the area of machine-learning. In contrast, as already noted above, neither the current claims nor the original disclosure even contemplated—much less actually implemented—any advanced technique for training a machine-learning or AI model. Instead, the specification already admits that the disclosed system/method is relying merely on existing training techniques, including: supervised learning, unsupervised learning, etc. (e.g., see [0022] to [0024] of the original specification). Consequently, Applicant’s attempt to show an alleged technological improvement, while repeatedly misapplying the two decisions above, is once again not persuasive.
Moreover, the Office has already established that Applicant’s claimed and disclosed technology is directed to the conventional computer/network technology. In particular, when each of the claims is considered as a whole, each claim is directed to the conventional and generic arrangement of the additional elements (also see above the discussion under Step 2B). Thus, the finding above confirms that the claims are indeed directed to well-understood, routine, conventional, activity (hereinafter WRCA) in the art. In contrast, while simply disregarding the Office’s finding above, Applicant is asserting that “[t]he Office Action has not established that this specific classification-driven, multi-stage processing architecture was well-understood, routine, or conventional”; and thus, Applicant’s assertion is inconsistent with the fact presented in the office action. It is also worth noting that the WRCA test is evaluating the claimed technology, but not the new abstract idea that the claims are reciting. For instance, the claims may overcome the prior art due to the new abstract idea that they are reciting; however, this does not necessarily mean that the claims are beyond the conventional computer/network technology. In this regard, Applicant appears to fail to properly apply the WRCA test.
Applicant also appears to be repeating the same arguments already addressed above. Applicant asserts “the claimed system produces a technical effect in the operation of the system itself, including improved structuring of prompting data, improved conditioning of inputs to the generative AI model, and reduction in redundant processing cycles caused by ineffective prompts. These effects demonstrate that the claim is directed to improving how the system operates, not merely to presenting information” (emphasis added).
However, except for relying on subjective theory, Applicant fails to demonstrate a factual technological improvement (if any) that the current claims (or the disclosure) are implementing. For instance, none of Applicant’s alleged “structuring of prompting data”, “conditioning of inputs to the generative AI model”, and “reduction in redundant processing cycles”, when considered individually or in any ordered combination, corresponds to a technological improvement. For instance, except for associating the term “improved” with the so-called “structuring of prompting data” and “conditioning of inputs”, Applicant fails to show a feature (if any)—or a combination of features (if any)—that supposedly provides the technological improvement. In fact, “structuring” and “conditioning” are very broad/generic terms; and therefore, none of these terms necessarily represents a technological feature, much less an advanced technological feature.
Similarly, Applicant also fails to demonstrate how the alleged “reduction in redundant processing cycles” is assumed to be a technological improvement. Given the context of the current claims and the original disclosure, the above appears to be merely an assumption regarding the number of interactions that the user is expected to make with the system. In particular, it is assuming that the user may not provide multiple redundant prompts since the system is providing the user with suggestions or guidance. However, this has nothing to do with a technological improvement. Instead, it is confirming the fact that the claimed system is being used—merely as a tool—to teach the user how to draft/construct a proper prompt. Moreover, while complying with the suggestions—or ignoring them—the user may continuously engage in dialogue with the system. This confirms that the system is operating/functioning in the same way regardless of the user’s decision. In particular, the system (a) collects input from the user, (b) analyzes the input, and (c) generates relevant information to the user. Thus, Applicant’s alleged “reduction in redundant processing cycles” does not seem to be materialized. Consequently, Applicant’s arguments fail to demonstrate whether any of the current claims (or the original disclosure), when considered as a whole, is implementing an element—or a combination of elements—that provides a technological improvement.
Accordingly, at least for the reasons above, the Office concludes that none of the current claims, when considered as a whole, implements an inventive concept that amounts to “significantly more” than an abstract idea.
Prior Art
5. Considering each of claims 1, 11 and 20 as a whole (including the respective dependent claims), the prior art does not teach or suggest the current claims (regarding the state of the prior art, see the office action dated 07/30/2025).
Conclusion
Applicant’s amendment necessitated the new grounds of rejection presented in this final 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 filled 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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRUK A GEBREMICHAEL whose telephone number is (571) 270-3079. The examiner can normally be reached from 7:00 AM - 3:00 PM.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, PETER VASAT can be reached on (571) 270-7625. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/BRUK A GEBREMICHAEL/Primary Examiner, Art Unit 3715
Prosecution Insights