Banks are increasingly adopting artificial intelligence (AI) to enhance their operations and customer experience. However, the challenge lies in scaling these AI capabilities across the organisation. There are four operating models for AI implementation in banking: centralised, distributed, hub and spoke, and centre of excellence.
The centralised model focuses on a single team handling all AI-related tasks, ensuring consistency but potentially slowing down implementation due to a lack of agility. The distributed model assigns AI tasks to individual business units, fostering innovation but risking a lack of coherence and standardisation.
The hub and spoke model combines the strengths of the centralised and distributed models. It has a central team that sets standards and provides support, whilst individual business units implement AI initiatives. This model balances standardisation with innovation but requires significant coordination.
The centre of excellence model has a dedicated team that develops best practices and provides training, but doesn’t implement AI initiatives itself. This model fosters expertise and consistency, but may lack the hands-on experience necessary for practical implementation.
Choosing the right model depends on the bank’s strategic objectives, existing capabilities, and organisational culture. Banks must also consider the trade-offs each model presents in terms of control, speed, and innovation.
Go to source article: https://www.mckinsey.com/industries/financial-services/our-insights/scaling-gen-ai-in-banking-choosing-the-best-operating-model