
A new report from UK Finance has shown that firms are benefitting from their investment in generative Artificial Intelligence (AI) and that ongoing collaboration with regulators, government and customers is important to support further innovation.
The report found forms are innovating with the technology but equally are being careful to manage the potential risks and ensure compliance with regulations.
The report therefore finds an emphasis on active human oversight for model training, interpretation and sensitive decision-making. The report stresses the importance of understanding each AI use case, ensuring high-quality data, and maintaining transparency in how models are used.
The report identifies three key risks linked to generative AI: firstly, reliability of outputs: generative AI models, particularly Large Language Models (LLMs), could produce bias, errors, or inappropriate language. Firms mitigate these risks by carefully selecting models, finetuning them using use case-specific datasets, and ongoing testing of outputs. Secondly is data privacy and security: in addition to risks common to many systems, such as inadequate handling of input data, LLMs can produce or reveal personal information in unexpected ways. Strong data protection practices and cybersecurity measures are critical, with new measures emerging such as personal information ‘filters’. Finally, third party considerations: Relying on external AI providers can reduce control. Firms can enhance third-party risk management processes to deal with this.
Overall, firms are benefitting from their investment in generative AI. By making use of their previous experience in safely using technology, they can adopt generative AI safely.
UK Finance says that to support further innovation, there needs to be ongoing collaboration with regulators and government and engagement with customers. Building trust through transparency and education, and ethical practices is key to ensuring AI’s potential is realised.
Jana Mackintosh, Managing Director of Payments and Innovation at UK Finance, said “Generative AI has created a lot of interest among the public and policymakers. It is an exciting new technology that has real potential, but also brings potential risks that will need to be managed. The financial services sector is currently focused on areas that involve active human oversight and are taking a careful approach. The good news is the sector has a strong track record of innovating responsibly with new technologies, positioning it well to harness the potential of generative AI.”