Bank technology

All cultures have explicit and implicit biases. Preventing discrimination when large language models are involved starts with using high-quality human-curated data and text.

Most banks are starting their generative AI experiments with closed-source models before progressing to open source. For some, the ultimate goal is to be model-agnostic.

Integrating artificial intelligence into banking operations needs to be done with a clear business purpose in mind, and the articulation of that purpose needs to come straight from the top of the organization.

There is fundamental work involved in laying a modern data foundation and adopting a flexible, data-driven, cloud-based IT architecture suitable for AI implementation. Most banks aren't there yet.

The financial services company has found that engineers report higher satisfaction and save time when using a copilot to assist with coding.

Branches won't succeed without significant investment in mobile and in-branch technology; but failing to invest in bankers themselves is just as big a problem.

The Dallas-based regional bank tapped a client for its co-pilot capabilities, where employees can message a bot instead of a human to get tech assistance.