AI Industry Innovations , AI Technologies , Banking & Finance

Gen AI Expected to Bring Big Changes to Financial Services

McKinsey Predicts Up to 15% Profit Increase, Gen AI Use in All Areas of Banking
Gen AI Expected to Bring Big Changes to Financial Services
Image: Shutterstock

While banks and financial services firms have primarily focused on enhancing productivity in their initial generative AI pilots, the technology holds the potential to significantly change job functions and customer interactions and pave the way for creating entirely new business models, according to McKinsey Global Institute.

See Also: Demo: Microsoft Copilot for Security

The industry has historically used AI applications to derive data-driven insights and foster agile decision-making. A McKinsey report found that the emergence of generative AI will significantly accelerate the maturity and overall impact of AI in banking. McKinsey predicted substantial opportunities in the banking sector, with an annual potential value gain ranging from $200 billion to $340 billion -equivalent to a 9% to 15% improvement in operating profit. McKinsey said this economic effect is expected to benefit all banking segments and functions, and the corporate and retail sectors are expected to experience the most substantial gains, reaching $56 billion and $54 billion, respectively.

Image: McKinsey & Co.

Fraud and Productivity Applications

Analysts expect three major areas to benefit from generative AI: customer service, risk management and operational efficiency. AI will also help banks, insurers and investment firms offer innovative financial solutions and advisory services, improve financial data analysis and reporting, and enhance collaboration and learning, McKinsey predicted.

In customer service, generative AI is changing the dynamics of client interaction for financial institutions. Through automated, personalized responses and enhanced advisory services, this technology is redefining the customer experience. Banks are increasingly adopting LLMs, including OpenAI's ChatGPT, to enrich customer engagement.

In terms of risk management, generative AI is improving fraud detection and risk assessment through its advanced analytics capabilities.

Generative AI also has a favorable effect on operational efficiency. It is accelerating claims processing in the insurance sector and streamlining loan approvals in banking. Accenture's research indicates that LLMs could affect up to 90% of working hours within the banking industry, potentially leading to a 30% increase in employee productivity by 2028. Generative AI is especially effective in automating compliance and underwriting processes, and it plays a pivotal role in these operational advancements.

Generative AI in Action

Beyond its theoretical potential, generative AI is being applied through real-world use cases across multiple functions.

  • Commonwealth Bank Australia: CBA uses generative AI in its call centers to assist staff in navigating complex customer queries by analyzing more than 4,500 policy documents in real time. Quick and accurate information helps staff address customer queries more effectively and efficiently, which enhances the overall customer experience.
  • HDFC ERGO: HDFC ERGO General Insurance has established a center of excellence for generative AI in collaboration with Google Cloud. This initiative focuses on creating hyper-personalized experiences, improving processes and driving cost efficiencies.
  • Nasdaq: Nasdaq is using generative AI in the areas of risk management and fraud detection for financial crime prevention and the development of AI-driven order types, including Dynamic M-ELO to enhance financial security.
  • JPMorgan Chase: The company filed a trademark application for IndexGPT, a ChatGPT-like LLM service for investment advice. This marks a significant stride in personalizing financial advisory and delivering customized financial solutions. It aims to offer customers tailored investment strategies by analyzing and selecting securities aligned to their specific needs.
  • Bloomberg: BloombergGPT, a 50-billion parameter LLM, is trained on financial documents curated by the company over the last four decades. The LLM aims to enhance data analysis and reporting in finance. This tool assists in complex tasks such as sentiment analysis, entity recognition and news classification, highlighting generative AI's prowess in handling specialized, domain-specific data.
  • Morgan Stanley Wealth Management: The firm has undertaken a strategic initiative with OpenAI to process and synthesize content to enhance knowledge management and collaborative learning. This initiative aids financial advisors in assimilating and processing extensive data and insights, thereby enhancing their ability to serve clients more effectively.
  • European Central Bank: The company is exploring the use of generative AI for various functions, including policy decision-making and document analysis. These use cases can be instrumental in enhancing the efficiency and accuracy of policy analysis and mitigating financial oversight.

As generative AI technology matures, its incorporation will become imperative for banking companies that are seeking a competitive edge and the ability to redefine industry benchmarks in service excellence, operational efficiency and innovation.

But adoption of gen AI in financial services comes with challenges - particularly cultural, ethical and operational ones. Businesses have to safeguard against the ongoing threat of AI hallucinations, in which generative AI models produce inaccurate and/or unintended output. Understanding what causes hallucinations and how to deal with them is crucial as the industry increasingly relies on AI and its responsible deployment.


About the Author

Shipra Malhotra

Shipra Malhotra

Managing Editor, ISMG

Malhotra has more than two decades of experience in technology journalism and public relations. She writes about enterprise technology and security-related issues and has worked at Biztech2.com, Dataquest and The Indian Express.




Around the Network

Our website uses cookies. Cookies enable us to provide the best experience possible and help us understand how visitors use our website. By browsing aitoday.io, you agree to our use of cookies.