Fazzaco Year-End AI Special Pt.II: Best 2025 AI Use Cases in Finance Explained

As artificial intelligence (AI) continues to gain momentum, more and more financial institutions in 2025 will feel the pressure to embrace this technology. However, instead of blindly following the trend, these institutions should focus on identifying the application scenarios that will deliver the greatest value to their businesses and governance.

A "Structured Approach"

Graham Smith, Head of Data Science and Innovation at NatWest Group, discussed in an interview with Finextra that converting pilots into actual offerings requires firms to adopt the right approach, processes, and management practices in order to achieve the maximum impact. He shared how NatWest has conducted extensive surveys to identify the best AI applications to help clients, shareholders, and businesses achieve a triple-win outcome.

As mentioned in the first article of our AI special​, Bahadir Yilmaz, Chief Analyst at ING Group, emphasized that financial institutions need to adopt a "structured approach" when determining how to integrate AI technology into their operations. This approach should align with risk management and compliance practices, prioritizing the ethical considerations of these use cases. For instance, at ING, every AI system undergoes 140 risk assessments before it is put into use.

"AI models based on personal data should be unbiased, explainable, transparent, responsible, and always grounded in consent. The use of data should not exceed its intended purpose, so clear regulations on data retention are essential. Currently, there are no specific, universal regulations regarding how AI models work. However, financial institutions are obligated to establish their own ethical frameworks for AI models," Yilmaz added.

Pavel Goldman-Kalaydin, Head of AI and Machine Learning at Sumsub, warned against the risk of blindly riding the latest AI trends. He suggested that institutions should focus on their specific pain points, combining AI with solutions to these challenges, rather than falling into the trap of "adopting AI just for the sake of it," in order to create real value.

Below, we explore several of the best AI use cases for 2025 as outlined in this research.

Use Case 1: Enhancing Customer Service Experience and Efficiency

In the financial services sector, especially for brokers dealing with retail traders, the primary use case of AI is to enhance the customer experience. As a result, AI chatbots are being integrated into businesses to simulate human interactions, significantly improving customer service efficiency. This technology not only optimizes customer communication but also paves the way for personalized services. With AI, financial institutions can predict customer needs, providing more streamlined and engaging interactions.

As AI has continued to evolve in 2024, the level of bespoke offerings in trading will reach unprecedented levels by 2025. Future trading services will be entirely AI-driven, offering tailored plans that align precisely with each customer's trading goals.

A case highlighted in the Finextra report refers to NatWest's AI chatbot, Cora, which has been in use since June this year. Since its launch, customer satisfaction has increased by 150%, and the number of customer issues requiring human intervention has been reduced by 50%.

Use Case 2: Combatting Financial Crime

As highlighted in our previous Fazzaco Whistleblower story, industry vermins like B2B Hub and Amiran Azaladze​ cause significant disruption and loss to brokers, not to mention the numerous traders falling victim to scams such as pig-butchering and Ponzi schemes. Therefore, tackling financial crime remains a top priority for financial institutions to address collaboratively.

By combining AI with other technologies while maintaining a human-centered approach, we now have the best opportunity to tackle this pressing challenge. AI can analyze vast amounts of real-time data, detect suspicious patterns, and ultimately help prevent fraud and identify new cyber threats.

Traditional fraud prevention methods often lead to false positives, flagging legitimate transactions as suspicious and inconveniencing traders. However, AI, by learning each user's distinctive behavior, can tailor transaction patterns to individuals. This personalized approach significantly reduces false positive rates, enhances the customer experience, and saves brokers time spent handling false alerts.

As financial criminals adopt increasingly sophisticated communication strategies, natural language processing (NLP) technology plays a key role in analyzing behavioral patterns. AI can also uncover hidden fraud networks. By performing graph analysis, AI can map relationships and identify potential links between seemingly independent behaviors, revealing large-scale fraud operations. This technological breakthrough enables systems to uncover complex fraud networks that would otherwise be difficult for humans to detect.

Overall, AI applications not only improve fraud detection and prevention efficiency but also assist financial institutions' legal, risk, and compliance teams in coping with evolving regulatory demands and scrutiny.

Use Case 3: Improving Internal and External Efficiency

AI's applications extend far beyond customer experience and scam alert. With technological advancements, AI-driven workflows will enable systems to make independent decisions. For instance, in banking, AI can automate credit scoring and loan approvals, dramatically improving loan process efficiency and reducing approval times.

As AI becomes more widespread across various industries in the coming years, credit risk analysis will undergo a fundamental transformation. Financial institutions will be able to leverage a broader range of data sources, such as social behavior and interactions, to provide more accurate credit assessments.

In the area of operational efficiency, AI is playing a crucial role at multiple stages. With call summary tools, AI helps relationship managers save time, allowing them to focus more on interacting with clients. Additionally, AI is being applied in internal workflows to support HR, streamline daily queries, and enhance conversational experiences, further improving employee satisfaction and productivity.

AI is also supporting financial institutions in addressing sustainability challenges. For internal sustainability teams working on green investment classification, AI can process vast amounts of data, enabling institutions to manage green investments more efficiently and make more informed decisions. These use cases demonstrate that AI is not just a tool to enhance customer experience; it is a key to improving overall operational efficiency and decision-making quality within institutions.

Next: AI Regulation in 2025

As we can see, the best AI use cases in the financial services sector focus on enhancing customer experience, combating financial crime, optimizing internal workflows, and driving the industry toward greater personalization and intelligence. However, the widespread adoption of AI also comes with risks, particularly in terms of ethics, compliance, and regulation. Financial institutions need to adopt a structured approach to ensure that AI applications comply with regulatory standards and prioritize potential ethical concerns and social impacts.

In the follow-up episode, we will explore the current state of legal and regulatory oversight of how AI is being applied in the financial industry, as well as the regulatory outlook for 2025.