.png){kind=logo}
As the field of artificial intelligence keeps transforming the financial service sector, banks have begun to look into intelligent technologies both to enhance customer interaction, automate support activities, and create more responsive online experiences. In this emerging area, technology researcher Bhavya Sri Sanku has registered a copyright for the research work titled as “Reinforcement Learning of Adaptive De-escalation Strategy Selection in Banking IVR and Chatbot Systems” under the Copyright Rules of Canada with Copyright No.1251225 dated October 26, 2026. Registered with the Canadian Intellectual Property Office (CIPO) in the literature works category, the study investigates how the reinforcement learning may be used to foster adaptable customer-service strategies and to confront significant copyright, ownership and regulatory issues.
Current banking organizations have to deal with large amounts of customer interactions, IVR voice response and chatbot powered by AI. These systems process queries that relate to transactions, account services, and payment disputes, as well as, fraud alerts among other financial matters. It can also become an issue among customers because an automated system is based on the set of pre-programmed answers, which do not identify emotional distress or shifts in the conversation. In research by Sanku, reinforcement learning is a potential solution in choosing a more suitable de-escalation method that can allow banking systems to react more wisely to problematic customer interactions.
The study relates to reinforcement learning, a machine learning method whereby systems acquire decision-making behaviors by minimizing errors via feedback and reward signals. In the suggested conversational model, customer interactions are evaluated, taking into account indicators like sentiment, requests that are recurring, unresolved issues, and the history of conversations by an intelligent system. According to these observations, a reinforcement learning policy will pick the correct response strategy, which may imply supportive communication, simplifying explanations, offering calls, or escalation to a human representative. The structure demonstrates the use of adaptive decision-making to facilitate more customer-service experiences that were context-sensitive.
One of the clear features of the work by Bhavya Sri Sanku is the presence in the work of the analysis of the intellectual property issues within the framework of the introduction of reinforcement learning into banking. Though much of current literature on the topic often focuses on the technical performance of AI systems, this paper explores the application of copyright law in Canada to the elements that underpin intelligent conversational systems. They consist of software source code, dialogue scripts, voice prompts, training sets, reward functions, trained model parameters, and AI-generated text. The analysis of each of these aspects in isolation leads to the development of the research providing a systematic view of the interconnection between technological advancement and protection of intellectual property.
Issues of ownership and authorship in AI-assisted software development are also discussed in the study. Banks tend to contract with technology vendors, independent developers, and outside service providers to initiate automated customer-support platforms. The presence of such arrangements can cause ambiguity on the ownership of software, trained models, conversational content, and related datasets. According to the analysis conducted by Bhavya Sri Sanku, having a well-defined contractual agreement, documented human contributions, and licensing of the contract and secrecy are crucial points of contention to facilitate sound development and deployment. Another major contribution to the research is on the utilization of customer interaction information to train intelligent systems. Sensitive financial information, personal information, and possibly copyrighted material can be found in banking conversations. This paper focuses on the regulatory issues of utilizing internal communication logs, licensed data, artificial dialogues, and third-party data to train machine learning.
The study also takes into account the general regulatory climate of artificial intelligence within Canadian financial institutions. The work determines considerations for organizations adopting reinforcement learning in systems facing the customer, through the analysis of copyright legislation, relevant judicial rulings in Canada, and guidance from the financial sector. Its advice recommends keeping records of training-data origin, ensuring the ownership of intellectual property, recording human authorship, handling automated responses to check authenticity and providing protection limitations on confidential model parts. This study has the potential to be applicable to more than traditional banking customer support. It also has potential to develop intelligent conversational systems based on adaptive de-escalation techniques in insurance, financial technology, telecommunications, and other service-based sectors where customer communication, operational reliability, and regulatory responsibility are critical.
In the case of Bhavya Sri Sanku, the Canadian copyright registration is a recorded breakthrough in the literature on the research of artificial intelligence, adaptive conversational systems and intellectual property regulation. The piece is a legal and conceptual analysis, not a report on some experimental performance outcomes. It provides a guideline on how intelligent customer-service technologies might be designed, keeping in mind the notion of ownership, compliance, and accountability. For financial entities still seeking AI-enhanced automation opportunities, studies of this type come into the conversation about creating a digital banking ecosystem that is more responsive, responsible, and legally aware.