The opinions expressed here are those of the authors. They do not necessarily reflect the views or positions of UK Finance or its members.

Digitisation through API integration, AI and Machine Learning have created new avenues for the modernisation of banking platforms. Historically, AI was predominantly used by a selective group of organisations like FinTechs and Digital Banks. 

However, the emergence of GenAI presents an opportunity for banks and financial institutions to leverage this cutting-edge technology. GenAI is simpler, easier and it requires less resources than traditional AI. With AI, you need modelers, data scientists, and a large amount of data. This is not something every institution can afford or have the resources and time to embrace it. GenAI is an opportunity for these institutions to jump into the wagon and tap into its benefits.

As GenAI has unlocked significant potential in the financial sector, fueled by increased curiosity and creativity, it also yielded critical lessons. A key learning point has been that there are numerous ethical implications.

• The first that comes to mind is bias and discrimination. GenAI models can lead to unfair outcomes if they are trained on biased data. Addressing biases in AI models is essential to ensure fairness and prevent discrimination, while fostering transparency, accountability and traceability. 
• Privacy is also critical. Misusing AI-generated content, and over-reliance on AI could potentially affect privacy breaches (think deepfake), it could also expose sensitive data in unexpected ways. Strong guardrails are required to maintain the privacy and accuracy of GenAI generated content.  
• Additionally, long term environmental and social impacts of AI must be considered, with attention to the creation of fake content and intellectual property rights. 

Maintaining human oversight is critical to ensure responsible AI deployment

To ensure that our GenAI models adhere to ethical principles, we have established a robust governance framework. This framework incorporates ethical guidelines into all stages of the AI development process, from design to deployment. By focusing on fairness, accountability, and transparency, we aim to mitigate potential risks and ensure that our AI systems are used ethically. 

To protect user privacy and data security, we adhere to relevant data privacy regulations and have implemented strong security measures. Our employees also play a crucial role in monitoring AI outputs and capturing user feedback to identify and address any issues early on. 

Furthermore, we leverage our partners' technologies and have established a secure architecture to safeguard against malicious use of GenAI. We leverage trusted sources of information, especially our Moody’s content, to drive high quality output. Our training programs for employees raise awareness about potential risks and promote responsible AI practices.

Banking adoption of GenAI: Other main considerations

A recent customer survey we conducted highlights the slow adoption of GenAI within medium and small-sized banks. While larger institutions have the resources to invest in GenAI, i.e. talent and infrastructure, smaller banks often face significant challenges. 

• One major concern is the regulatory environment. Banks operate under strict regulations, and the introduction of new technologies like GenAI requires careful consideration and approval.
• Another key concern is data privacy and security. Banks handle vast amounts of sensitive customer data, which must be protected from unauthorised access, use, or disclosure. Ensuring that GenAI models can process this data safely and securely is a critical challenge.
• Transparency and explainability is also a major concern. Many GenAI models are considered "black boxes," making it difficult to understand how they arrive at their decisions. Regulators often demand transparency and explainability, which can be challenging to achieve with complex AI models.
• Finally, introducing new technologies can be disruptive and may face resistance from employees who are comfortable with existing processes. Overcoming this resistance requires careful change management and communication to ensure a smooth transition. 

Moody's has developed a comprehensive infrastructure and governance process to address all key concerns banks face when adopting GenAI. Our customers are our partners in this process. The position of trust allows our expertise to provide guidance and support to banks in navigating regulatory compliance. 

As we all adapt to this new technology – it’s important to talk about the success and the progress made. Several banks have successfully implemented our AI applications and realised significant benefits and value. It is exciting – being able to collaborate with industry practitioners on this transformative journey. 

Find out more about how Moody’s is unlocking the next generation of technology and putting it to work for banks here.