The integration of generative artificial intelligence into financial services has entered a new phase characterized by practical implementation rather than speculation.
This shift comes three years after the technology entered mainstream consciousness in late 2021, with companies now focusing on commercial viability rather than technical capabilities.
In Mastercard's latest Signals report, Microsoft founder Bill Gates calls the technology “the most important advancement since the graphical user interface,” although OpenAI CEO Sam Altman admits he's “a little bit scared” of the technology his company is developing has.
The financial sector's approach to generative AI implementation focuses on three key developments: informed AI, which combines large language models with external data sources; perceptual AI that interprets environmental data; and proactive AI that operates with reduced human supervision.
Informed use
Banks and payment providers implement informed AI through two main methods: fine-tuning and retrieval-augmented generation (RAG). Fine-tuning involves training an AI model on specific data sets, while RAG allows AI systems to access external databases in real time.
The choice between these methods depends on data quality, update requirements and security considerations.
This is particularly relevant in the financial and legal sectors, where protecting confidential information from unauthorized access is of utmost importance. Current data shows that 20% of companies use fine-tuning, while 80% use RAG to complement their language models.
In the banking sector, 73% of mortgage lenders see generative AI as central to improving operational efficiency in lending processes.
The technology allows underwriters to incorporate market conditions and real estate trends into their analyses. Financial institutions use the technology for fraud monitoring, transaction analysis and personalized financial advice.
Enterprise technology provider Cohere offers RAG capabilities tailored to business needs, while data processing company Unstructured provides technology to convert unstructured data into RAG-compatible formats.
Glean, a workplace technology provider, has developed an AI assistant that uses RAG to display relevant information for financial professionals.